Oxazepam.

Dependence, tolerance and withdrawal are the defining hazards and oxazepam is not exempt from them. A literature review of the benzodiazepine withdrawal syndrome documents reactions with oxazepam alongside chlordiazepoxide, diazepam, lorazepam and others, notes that risk rises with dose and duration, and records that withdrawal can occur even at recommended doses after short term therapy ^[6]. In a randomised comparison of benzodiazepine detoxification strategies, patients tapered on oxazepam experienced paradoxical symptoms not seen in the flumazenil arm and relapsed more often at days 15, 23 and 30 ^[10]. Long term users remain pharmacologically responsive rather than fully tolerant; in a double blind crossover study, patients taking oxazepam daily for more than 10 years still showed dose related effects on saccadic eye movements and startle response from an extra dose, and memory impairment did not differ from benzodiazepine naive controls ^[11]. Driving impairment is real and steeply dose dependent. In an on the road highway driving test, mean standard deviation of lateral position rose by 1.83 cm after oxazepam 10 mg and by 7.57 cm after oxazepam 30 mg, the latter classed as severe impairment and larger than diazepam 10 mg at 3.03 cm ^[5]. Sedation, drowsiness, dizziness, ataxia and psychomotor slowing follow from the same mechanism. Anterograde and implicit memory effects appear milder than with lorazepam: at all test intervals, lorazepam 2.5 mg impaired perceptual priming while oxazepam 30 mg did not ^[12], and in a two week double blind study oxazepam 15 mg three times daily behaved like placebo on immediate memory while nordiazepam and chlordiazepoxide lactam impaired it ^[13]. In elderly people the risk is injury rather than only sedation; in a five year prospective cohort of 253,244 new elderly users, higher doses of oxazepam, flurazepam and chlordiazepoxide carried the greatest injury risk, and risk varied by drug independently of half life ^[14]. Absorption is slow and formulation dependent, and elimination half life was significantly prolonged in elderly subjects ^[15]. Liver disease is the one place oxazepam looks better than its class: clearance, half life, volume of distribution and protein binding were all comparable to age matched controls in acute viral hepatitis and cirrhosis ^[1], and in vitro oxazepam glucuronidation was unchanged in microsomes from cirrhotic livers while zidovudine and lidocaine clearance fell ^[16]. That protection is not unlimited; severe decompensated liver disease did reduce oxazepam clearance, and hypothyroidism and very low calorie dieting reduced it as well ^[17]. Interactions are comparatively few because phase I metabolism is bypassed: cimetidine prolonged the half life of desmethyldiazepam but left oxazepam disposition unaffected ^[18]. Additive central nervous system depression with alcohol, opioids, sedating antihistamines and other depressants remains the dangerous combination, and combining oxazepam with alcohol retarded learning acquisition where oxazepam alone did not ^[13]. Aggression and disinhibition are less prominent than with chlordiazepoxide but not absent; in human hostility studies oxazepam did not substantially disinhibit hostility while chlordiazepoxide did ^[19], yet in a laboratory aggression paradigm in men, oxazepam has been examined alongside diazepam and clorazepate in the same class of paradigm ^[20], and in rats oxazepam at 2.5 to 80 mg per kilogram induced mouse killing as potently as chlordiazepoxide ^[21]. Rodent carcinogenicity is a genuine finding that requires proportion; in two year National Toxicology Program feed studies there was equivocal evidence of carcinogenic activity in male F344/N rats based on small increases in renal tubule adenoma against markedly enhanced nephropathy, no evidence in females, and oxazepam was not mutagenic in Salmonella, did not induce chromosomal aberrations, and was negative in the mouse micronucleus test ^[22]. The mouse liver tumours reported at high dietary doses involve Wnt and beta-catenin signalling and oxidative stress in animals given far above clinical exposures ^[23], and mechanistic work places oxazepam among CYP2B and CYP4A inducers in mice rather than genotoxic agents ^[24]. Abuse liability exists but sits low within the class; diazepam was 2.6 to 5.7 times more potent than oxazepam across psychomotor, cognitive and subjective measures and was relatively more potent still in producing liking, with a faster onset of 1 to 2 hours against 4 to 12 hours for oxazepam ^[4]. In pregnancy, benzodiazepine exposure in the first trimester has not been convincingly associated with major malformations, though the literature carries recall bias and confounding ^[25]; benzodiazepine use late in pregnancy and during breastfeeding remains a caution. Oxazepam and its parent drugs can remain detectable for far longer than expected, with diazepam metabolites found in urine 79 days after withdrawal in one documented case ^[26].

Oxazepam was developed at Wyeth as Wy-3498 and reached clinical use in the mid 1960s, with regulatory approval in 1965. It arrived as a deliberately simpler molecule than the first generation benzodiazepines: chemists had already established that chlordiazepoxide and diazepam are converted in the body to hydroxylated products, and oxazepam is that endpoint given directly. Controlled trials in psychoneurotic inpatients and in office internal medicine practice were run in 1966 ^[3]. It was later marketed worldwide under names including Serax, Tazepam, Seresta, Sobril and Praxiten, and became a mainstay in Scandinavian and continental European alcohol withdrawal protocols. Its role in the literature widened again in the 2000s when it was paired with metyrapone as an investigational combination for cocaine use disorder.

Oxazepam has a reputation as the mild, forgiving benzodiazepine: slow to come on, free of active metabolites, unusually safe in liver disease and old age, and less liked by people who use drugs recreationally. Clinicians who prefer it cite the glucuronidation pathway and the absence of accumulation; a nursing pharmacology review went so far as to ask whether it is underutilised ^[9]. That gentle reputation is broadly supported but has been oversold in two places, since a 30 mg dose caused severe on the road driving impairment ^[5] and higher doses ranked among the highest injury risks in elderly users ^[14]. In continental Europe it remains a standard alcohol withdrawal agent; in the United States it is now a quieter, older option than alprazolam or lorazepam.