DRUG DISPOSITION

nino Pharmacokinel. 23 (6): 415-427. 1992 0312-5963/92/0012-0415/$06.50/0 © Adis International Limited. All rights reserved. CPK1240

Clinical Pharmacokinetics of Ketorolac Tromethamine Dian R. Bracks and Fakhreddin Jama/i Faculty of Pharmacy and Pharmaceutical Sciences. University of Alberta. Edmonton. Alberta. Canada

Contents 415 416 417 417

419 420

421 423 423 424

424 425 425 425 426

Summary

Summary I. Analytical Methods 2. Pharmacokinetic Properties 2.1 Absorption 2.2 Distribution 2.3 Elimination 3. Stereoselective Pharmacokinetics 4. Implications of Pharmacokinetic Properties for Therapeutic Use 4.1 Dosage and Therapeutic Range 4.2 Influence of Age on Ketorolac Pharmacokinetics 4.3 Effects of Renal and Hepatic Diseases 5. Pharmacokinetic Drug Interactions 5.1 Effect of Other Drugs on the Pharmacokinetics of Ketorolac 5.2 Effect of Ketorolac on the Pharmacokinetics of Other Drugs 6. Conclusions

Ketorolac is a new chiral nonsteroidal anti-inflammatory drug (NSAID) which is marketed for analgesia as the racemate. The drug is administered as the water soluble tromethamine salt and is available in tablets or as an intramuscular injection. The absorption of ketorolac is rapid, C max being attained between 20 to 60 min. Its oral bioavailability is estimated to range from 80 to 100%. The drug is extensively bound (>99%) to plasma proteins and has a volume of distribution (0.1 to 0.3 L/kg) comparable with those of other NSAIDs. Only small concentrations of ketorolac are detectable in umbilical vein blood after administration to women in labour. The elimination half-life is between 4 and 6h and is moderate in comparison with other NSAIDs. The area under the plasma concentration-time curve of ketorolac is proportional to the dose after intramuscular administration of therapeutic doses to young healthy volunteers. Ketorolac is extensively metabolised through glucuronidation and oxidation; little if any drug is eliminated unchanged. Most of the dose of ketorolac is recovered in the urine as conjugated drug. Although ketorolac is excreted into the breast milk, the amount of drug transferred comprises only a small fraction of the maternal exposure. Little stereoselectivity was present in the pharmacokinetics of ketorolac in a healthy volunteer receiving single intravenous or oral doses. The elderly exhibit reduced clearance of the drug. Renal insufficiency appears to cause an accumulation of ketorolac in plasma, although hepatic disease may not affect the pharmacokinetics.

Clin. Pharmacokinet. 23 (6) 1992

416

Ketorolac (fig. 1) is a new racemic nonsteroidal anti-inflammatory drug (NSAID) which is marketed for systemic use as a peripherally acting analgesic agent (Rooks et al. 1982, 1985). The drug was recently introduced into the pharmaceutical markets of several countries, including the US, Canada, New Zealand, Italy and Denmark. The analgesic properties of ketorolac are manifested through its prostaglandin synthetase inhibitory activity (Rooks etal. 1982, 1985) which, like other aryl alkanoic acid NSAIDs, is mostly attributable to the S enantiomeric configuration (Guzman et al. 1986; Jamali 1988). Because ketorolac is produced as a tromethamine salt, it possesses sufficient water solubility to allow for its parenteral administration (Rooks 1990). Other NSAIDs which have also been used parenterally are diclofenac sodium (Reynolds 1989), indomethacin sodium trihydrate (Reynolds 1989) and ketoprofen (Debruyne et al. 1987). Ketorolac seems to be as effective as morphine, but is without potentially troublesome side effects such as respiratory depression or constipation (Resman-Targoff 1990). This may allow ketorolac a useful role in the short term alleviation of postoperative pain in surgical patients (Buckley & Brogden 1990; Forbes et al. 1990; Resman-Targoff 1990; Stanski et al. 1990). Because ketorolac is an NSAID, however, it does share many of the same side effects as other NSAIDs (Buckley & Brogden 1990; Litvak & McEvoy 1990). It has also been studied for use as an anti-inflammatory agent in the treatment of ocular disorders (Aach et al. 1988). Several general review articles are available which deal with the pharmacodynamics, pharmacokinetics and therapeutic uses of ketorolac

Cl \" %~

o II

*COH

",,;

C

N

II

o Fig. 1. Structure ofketorolac tromethamine. centre.

* Denotes chiral

5.00

... Intravenous • Intramuscular • Oral

::J'

0; 1.00

.s .§

:8

0.10

8 0.Q1

+-~---r-~.--""'---'---'~'-~"""""-""""""

o

4

8

12 Time (h)

16

20

24

Fig.2. Mean plasma concentration-time profiles ofketorolac in 15 healthy volunteers given intravenous, intramuscular and oral doses of ketorolac tromethamine IOmg (from Jung et al. 1988, with permission).

(Buckley & Brogden 1990; Litvak & McEvoy 1990; Resman-Targoff 1990). In this article we review the clinical pharmacokinetics of ketorolac. The pharmacokinetics of the enantiomers in a healthy volunteer are also reported for the first time.

1. A.nalytical Methods Very few assay procedures have been published for the quantitative analysis of ketorolac in biological specimens (Jamali et al. 1989b; Wu et al. 1986). The manufacturer, Syntex, has developed an assay for its quantification, although it has only been published as an abstract (Wu et al. 1986). Mroszczak et al. (1987) briefly describe another method, but details were not included. All of the available methods involve initial acidification of the sample, followed by extraction into an organic solvent. Ultraviolet absorbance is used for detection by all methods, either at 280 or 313nm. Radiolabelled ketorolac has also been used to follow the time course of the drug and its metabolites in biological samples (Ling & Combs 1987; Mroszczak et al. 1987). The assay of Wu et al. (1986) is also capable of measuring an inactive metabolite, p-hydroxy-ketorolac, in human plasma (Jung et al. 1989). To date there is only 1 assay procedure, re-

Pharmacokinetics of Ketorolac

ported in detail, for the separate quantification of the enantiomers ofketorolac (Jamali et al. 1989b). This is of importance because it is the S-enantiomer which gives ketorolac most of its analgesic activity.

2. Pharmacokinetic Properties 2.1 Absorption When administered orally or intramuscularly as the tromethamine salt, ketorolac is rapidly and well absorbed. The maximal plasma concentration (C max ) is attained between 20 to 60 min (table I), which is a useful property for an analgesic drug in which a short onset of action is most desirable. In 2 studies, both oral and intravenous matching doses of ketorolac tromethamine were administered to healthy volunteers. In these studies the absolute bioavailability was 80.5 ± 9.6% following 1.7 mgf kg doses of an oral solution to 4 volunteers (Mroszczak et al. 1987) and 100 ± 19.8% following the administration of lOmg tablets to 15 volunteers (Jung et al. 1988; fig. 2); all volunteers fasted the night before the study. The discrepancy between the estimates of bioavailability (F) derived from these 2 studies was addressed by Jung et al. (1988), who suggested that Mroszczak et al. (1987) enrolled too few participants (n = 4) and that the variation was too great to allow for an estimation of F. However, the coefficient of variation in the F determined by Mroszczak et al. (1987) was lower (11.9%) than the 19.8% obtained in the study by Jung et al. (1988). The latter investigators also postulated that some of the oral solution was degraded in the gastrointestinal tract before being absorbed. This speculation is disputed by a published abstract (Mroszczak et al. 1985), in which oral solutions, tablets and capsules of ketorolac tromethamine were stated to be bioequi valent. Furthermore, because drugs in solution are generally absorbed faster than tablets, less time may be available for degradation to occur after solution administration. Therefore, we cannot discount the value of F cited by Mroszczak et al. (1987). Another potential cause of the discrepancy not addressed by Jung et al (1988) is the possibility

417

of differences in the assay methods used. It may also be pertinent that the dose in the study by Mroszczak et al. (1987) was considerably greater than that administered by J ung et al. (1988). The highest oral do~ previously studied was 234mg, although pharmacokinetic details were not provided (Mroszczak et al. 1985). In a review paper, Mroszczak et al. (1990) described the results of a study by Syntex in which oral doses of ketorolac tromethamine 10mg were administered to 12 healthy young volunteers. The lag time for absorption was 0.2h after administration and the absorption half-life (t'l2a) was estimated as 3.8 min. The same review reported a study of the effects of food on ketorolac absorption conducted by the manufacturers. Compared with ketorolac tromethamine 10mg under fasting conditions in the same 12 healthy young volunteers, the absorption of the drug was delayed by the ingestion of a high-fat breakfast, as evidenced by significantly lower e max and longer time to C max (tmax) [fig. 3]. There was no difference in the total amount absorbed (estimated from the AUC) or the elimination half-life (t'l2fj). Intramuscular ketorolac appears to have the same bioavailability as intravenous doses (Jung et al. 1988). The bioavailability of oral ketorolac rel-

• Fasting A AntaCid • High-fat breakfast

1°75 8 0.50

I

n:

0.25

2

4

6

8 10 12

24

Time (h) Fig. 3. Effects of a high-fat breakfast and an aluminum/mag-

nesium hydroxide-containing antacid on the mean plasma concentration-time profile of ketorolac (from Mroszczak et al. 1990, with permission).

418

Clin. Pharmacokinet. 23 (6) 1992

Table I. Mean pharmacokinetic parameters of ketorolac when administered to humans in single doses of the tromethamine salt Subjects

n

Mean age

Mean

Dosea

t max

(years)

body-

(mg)

(min)

C max (mg/L)

72

10 PO

53

[range]

References

AUC

CL

Vd

(mg/ L· h)

(L/h/kg)

(L/kg)

t'l2ll (h)

(%)

0.81

4.81

0.020

0.15 b

5.07

1()()C

46

0.77 0.86

0.018 0.021 0.033

0.13b 0.17 0.22b

4.99 5.09 4.69

98 b

20

5.19 4.82 2.84

75 M

Jallad et al. (1990)

weight

F

(kg) [range] Healthy young volunteers

15

32 (22-40)

Jung et al. (1988)

Healthy young volunteers

16

30 (20-39)

75.2

101M 10lV 10 PO

Healthy young volunteers

12

29.9 (24-38)

80.7

301M 30 PO

45 30

2.99 2.70

11.3 12.5

0.026 0.020

O.17 b 0.16

4.45 5.56

91 M

Jung et al. (1989)

301M 60 1M 90 1M 100-145 POb.e

50 48 45 30

2.24 4.48 6.88 10.1

13.7 26.1 40.7 28.8

0,018 0.019 0.019

0.18 b 0.15 b 0.15b

5.21 5.42 5.52 6.0

81 C

Mroszczak et al. (1987)

35.6

0.033

Healthy young volunteers

4

(27-35)

(59-85)

4

(27-35)

(59-85)

8

(23-35)

(66-76)

0.25 b

Healthy young volunteers

16

30

100-145 IVb 86-99 1Mb 10 PO

Healthy young volunteers

16

28

301M 10 PO

45 20

2.99 0.87

11.3 2.84

0.026 0.037

O.17 b 0.25 b

4.45 4.69

35

2.99 0.86

11.3

12

301M 10 PO

45

Healthy volunteers

0.026 0.027

O.17 b 0.21

4.45 5.41

Hepatic impairment

7

51

10 PO

46

0.87

3.23

0.032

0.21b

4.46

76M

10

57

301M 10 PO

37

Renal impairment

43

2.62 0.92

12.7 7.90

0.029 0.019

0.23b 0.27 b

5.43 9.91

94 M

2.57

10

6.0 (4.3-8.1)

20 (16-24)

301M 8.1-12 IVb

50

Paediatric pts (postsurgical)

25.1 11.9b

0.016 0.042b

0.22b 0.36b

9.62 6.1

Healthy elderly volunteers

13

72.1 (65-78)

74.5

10 PO

43

0.90

4.04

0.023

0.20b

6.14

301M

58

2.52

15.3

0.019

0.19b

6.95

5.2

45

5.9

29.5

20

0.86

2.84

0.033

0.22b

4.69

75 M

75b.d

5.0

Mroszczak et al. (1985) Pages et al. (1987) Martinez et al. (1987) Olkkola & Maunuksela (1991)

79 b•d

= 0.678). Oral doses given as tablets unless otherwise specified.

a

As the tromethamine salt (conversion

b

Calculated by the present authors from data reported in the original paper.

c

Compared with matching intravenous doses.

d

Compared with matching intramuscular doses.

e

Given as solution. Abbreviations: C max peak plasma concentration; t max time to Cmax ; AUC area under the plasma concentration-time curve; CL total body clearance, expressed as CL/F for oral doses; Vd volume of distribution, expressed as Vd/F for oral doses; elimination half-life; F systemic bioavailability; 1M intramuscular; IV intravenous; PO oral. t'l2ll

= =

=

=

=

=

=

=

=

419

Pharmacokinetics of Ketorolac

ative to intramuscular doses ranges from 75 to 91 % in healthy young volunteers (table I). 2.2 Distribution

2.2.1 Protein Binding The mean volume of distribution (Vd) of ketorolac in healthy volunteers ranges from 0.13 to 0.25 L/kg (table I). Similar to other NSAIDs with related chemical structures (Lin et al. 1987), ketorolac is extensively bound to plasma proteins. In human plasma spiked with concentrations of racemic [14C]ketorolac tromethamine ranging from 0.5 to 10 mg/L, the mean percentage of drug bound ranged from 99.1 to 99.3% (Mroszczak et al. 1987) using equilibrium dialysis of plasma vs phosphate buffer (pH 7.4) at 3rc. These findings were later corroborated by lallad et al. (1990), who used equilibrium dialysis and [14C]ketorolac to determine the plasma protein binding in samples collected 1 or 4h after administering intramuscular ketorolac tromethamine 30mg to volunteers. In young . individuals, at 1 and 4h postdose the percentage of ketorolac bound to plasma was 99.1 and 99.3%, respectively. The corresponding mean plasma concentrations were approximately 2.5 mg/L at Ih and 1 myL at 4h. Interestingly, tht< mean AVC of oral solutions in the volunteers studied by Mroszczak et al. (1987) was only 4.6-fold greater than that reported by lung et al. (1988) after tablets, despite a 10- to 14.5-fold greater dose in the former study (table I). A discrepancy also occurred with the intravenous doses, in which the same difference in dose was present, but the AVC was only 7.4-fold higher after the larger doses. This apparent deviation from linearity, which was not addressed by lung et al. (1988), may perhaps be explained on the basis of nonlinear plasma protein binding at higher plasma concentrations, with a compensatory increase in total body clearance (CL). As shown in table I, in the study by Mroszczak et al. (1987) CL was 57% greater than that determined by lung et al. (1988). Saturable protein binding and enhanced CL of another NSAID, naproxen, also occurs with increasing plasma concentrations (Runkel et al. 1974).

0.3 0.25 .Q

~

•• •

0.2

> 0.15 :::E > =>

•

• • .".•

...- ••

0.1

• •: •

0.05

~

•

• •

• ••

•

0 0

2

3

4

5

6

7

Time (h)

Fig. 4. Umbilical vein to maternal vein ratios of plasma concentrations (UV: MV ratio) ofketorolac tromethamine in 3 mothers given ketorolac tromethamine IOmg. Doses were given during labour and concentrations were obtained after delivery (from Walker et al. 1988, with permission).

The relative plasma protein binding of ketorolac enantiomers has not been reported.

2.2.2 Distribution into Fetal Blood Ketorolac has been studied as an analgesic for relieving the pain of labour because of its low incidence of sedation and respiratory depression compared with narcotic analgesics. Consequently, ketorolac concentrations were studied in fetal blood to determine the amount transferred from maternal blood (Walker et al. 1988). 32 women (mean 27 years) were given single intramuscular ketorolac tromethamine 10mg doses during labour and blood samples were obtained after delivery from the mother and from the umbilical cord. The sampling times ranged from 0.57 to 6.57h after administration. The umbilical cord plasma concentrations were 0.017 to 0.119 myL, whereas corresponding maternal plasma concentrations were 0.223 to 0.873 mg/L. There was relatively little ketorolac transferred from maternal to fetal blood, as evidenced by a mean umbilical: maternal plasma concentration ratio of 0.116 (fig. 4) which appeared to increase with time after administration. As stated by Walker et al. (1988), in this respect ketorolac has an advantage over some narcotic analgesics, for which this ratio is 0.61 for pethidine (meperidine) [Kanto 1986] and even greater for nalbuphine (Wilson et al. 1986).

420

2.2.3 Distribution Into Other Tissues Although the distribution of ketorolac to other human tissues is not known, its distribution has been studied in tissues of Swiss-Webster albino mice following oral doses of [14C]ketorolac (Mroszczak et al. 1987). The AUC which we calculated using the linear trapezoidal rule [expressed in parentheses as ~g equivalents/(ml or g) • h) was highest in the kidney (33), followed by plasma (22), liver (4.7) and lung (4.6). Lower concentrations were present in muscle (2.2), gonads (2.3) and spleen (2.1). The lowest concentrations were in fat (1.7) and brain (0.72) tissues. The high concentrations in the kidneys might be attributed to urinary excretion of ketorolac in the mouse, which is the primary route of elimination of the parent drug and its metabolites in that species. This finding may be clinically relevant, because of the importance of renal excretion in the elimination of ketorolac and its metabolites in humans. With its potential as a useful anti-inflammatory agent for the eye (Aach et al. 1988), the distribution of [14C]ketorolac to ocular tissues was investigated in the rabbit (Ling & Combs 1987). A total of3.3% of the dose was recovered from the various eye tissues and fluids 24h after single topical application of 50~L of a 0.5% ketorolac tromethamine solution to the eyes. The highest concentrations were seen in scleral and corneal tissues. Small concentrations of [14C]ketorolac equivalents were detected in some of the ocular tissues and fluids examined after a Img intravenous dose. 2.3 Elimination Ketorolac has a t'/21l (4 to 6h; table I) which is of moderate length in comparison with that of other NSAIDs (Verbeeck et al. 1983). Mroszczak et al. (1987) found 92% of radio labelled ketorolac in the excreted urine of healthy volunteers within 24h postdose. Although up to 60% of the dose was found in urine as unchanged drug, the investigators suggested that the true amount excreted unchanged is between 5 and 10%. The higher amount detected in the samples was attributed to spontaneous hydrolysis of the labile acylglucuronides, which may

Clin. Pharmacokinet. 23 (6) 1992

50

40

s:

--

...J

Cl

~

()

:::l

30

0(

c:

os Q)

~

20

10

o

10

20

30

40

50

60

70

Dose (mg) Fig. 5. Relationship between dose and mean area under the plasma concentration-time curve (AUC) of ketorolac in healthy volunteers receiving intramuscular ketorolac tromethamine 30, 60 or 90mg (from Jung et al. 1989, with permission)

occur with NSAIDs during sample collection and storage (Upton et al. 1980). Only 5.9 to 6.3% of the radiolabelled dose was recovered in faeces, suggesting either limited biliary excretion or extensive enterohepatic recirculation. The pharmacokinetics of ketorolac are linear within the normal range of therapeutic dosages. Following single intramuscular doses of ketorolac tromethamine of 30, 60 and 90mg to healthy volunteers (fig. 5), the AVC of ketorolac increased linearly (Jung et al. 1989), with no change in the t'/21l. This was also reported to occur with oral doses, although details were not provided (Mroszczak et al. 1985). A deviation from linearity (increased CL) following doses larger than those recommended is plausible, although not clearly addressed in the literature (see section 2.2.1).

421

Pharmacokinetics of Ketorolac

There is no published study describing the pharmacokinetics of ketorolac after administration of multiple doses. However, a published abstract (Mroszczak et al. 1985) states that steady-state concentrations are attained within 24h when the drug is administered 4 times daily. A review paper also states that the manufacturer has found no appreciable accumulation after multiple-dose administration of ketorolac tromethamine 10mg orally or 90mg intramuscularly (Buckley & Brogden 1990). Wischnik et al. (1989) studied the excretion of ketorolac into the breastmilk of 10 lactating mothers (age 27 ± 4.3 years) following multiple daily doses of 10mg 4 times daily for 4 days. In women in whom the breastmilk concentrations were >5 !-(g/L, the milk: plasma concentration ratios were very low, ranging from 0.015 to 0.028 after 2h postdose. The low concentrations in breastmilk were attributed to the high degree of plasma protein binding of ketorolac and the lower pH of breastmilk than plasma. The investigators calculated the maximal possible infant exposure to ketorolac via breastmilk to be 0.4% of maternal exposure. This compares favourably with other NSAIDs such as diclofenac, ibuprofen and naproxen, in which infant exposures were 1.2% (Todd & Sorkin 1988), 0.8% (Townsend et al. 1984) and 0.26% (Jamali & Stevens 1983), of the corresponding maternal exposures, respectively. After administration of the commonly used analgesic paracetamol (acetaminophen), suckling infants are exposed to 1.85% of the maternal exposure (Notarianni et al. 1987). 2.3.1 Metabolism Ketorolac is extensively metabolised to inactive or mainly inactive metabolites. The major metabolite of ketorolac, the acyl-glucuronide, appears to account for approximately 72 to 77% of the amount of an oral dose excreted in urine (Mroszczak et al. 1987). The other major metabolite of ketorolac is p-hydroxy-ketorolac, which represents up to 12% of the amount of drug excreted in urine following oral doses ofketorolac tromethamine 100 to 145mg. p-Hydroxy-ketorolac appears to have 20% of the anti-inflammatory activity and 1% of the analgesic activity of ketorolac. As with the parent drug itself,

the pharmacokinetics of p-hydroxy-ketorolac are linear following single intramuscular doses of ketorolac tromethamine 30 to 90mg intramuscularly (Jung et al. 1989). p-Hydroxy-ketorolac is seen in plasma following intramuscular and oral administration of ketorolac tromethamine, and the AUC of the metabolite is similar after either route of administration, providing evidence that the metabolism of ketorolac after oral administration is similar to that following intramuscular use (Jung et al. 1989). Other unidentified metabolites in urine account for only 6 to 7% of the urinary recovery of [14C]ketorolac. The possibility of renal glucuronidation has been suggested in order to explain the apparently high urinary recovery of the acyl-glucuronide metabolite. This suggestion was based on in vitro studies using microsomes of renal origin (obtained from rabbits), which possessed 6-fold more activity than microsomes of hepatic origin (Mroszczak et al. 1987). Unfortunately, details of this experiment were not published. Considerable renal glucuronidation has been seen with another NSAID, ibuprofen, in the perfused rat kidney (Ahn et al. 1991). The extraction ratio of ketorolac is low and can be calculated to be between 0.002 to 0.09 using the estimates of bioavailability of oral ketorolac tromethamine provided by Mroszczak et al. (1987) and Jung et al. (1988), respectively. Therefore, the elimination of ketorolac is highly dependent on intrinsic clearance through metabolism of the unbound fraction in plasma, and not dependent on hepatic blood flow.

3. Stereoselective Pharmacokinetics Ketorolac is a 2-arylpropionic acid derivative and, similar to several other agents in this class, is marketed as the racemate. For ketorolac this is an important consideration, because it has been determined that the S-( -) enantiomer is much more active than the R-(+) enantiomer (Guzman et al. 1986). The S: R ratios of anti-inflammatory (rat carageenan oedema test) and analgesic (mouse phenylquinone writhing assay) activities are 57 and 230, respectively. Unlike other NSAIDs, the S en-

Clin. Pharmacokinet. 23 (6) 1992

422

aritiomer has the (-), rather than the (+) optical rotation. It is common for enantiomers of drugs to possess stereoselectivity in their pharmacokinetics (Jamali 1988; Jamali et al. 1989a). In reviewing the literature, it became apparent to us that the enantioselective pharmacokinetics of ketorolac have not been studied in humans. Therefore, for this review article, we administered single IOmg doses of racemic ketorolac tromethamine (Toradol®, Syntex Inc., Mississauga, Ontario, Canada) to a healthy human volunteer (age 45 years, bodyweight 74kg) on 2 separate occasions. On one day, a single IOmg tablet was taken, followed 2 weeks later by a single IOmg dose of the parenteral injectable formulation given intravenously. Blood samples were collected in heparinised tubes and plasma was separated by centrifugation. Urine was collected at intervals shown in figure 6. Plasma and urine were assayed using a previously described high pressure liquid chromatography technique (Jamali et al. I 989b), although the UV wavelength was set at 313 rather than 280nm. Part of each sample was hydrolysed

with NaOH to assess for the presence of glucuronide conjugates of ketorolac. The plasma concentration vs time profiles of the enantiomers are presented in figure 6. It is evident that the enantiomers possess very similar pharmacokinetics in the study participant (table II). The t'l2.8 was slightly shorter than reported by others for total (R,S)-ketorolac, although in our study the time course was only followed for 8h, which might have caused an underestimation of the true t'l2.8. No conjugated ketorolac was observed in the plasma samples, although after alkali hydrolysis, a substantial portion of the dose of each enantiomer was recovered in urine (fig. 6, table II). No evidence of chiral inversion was seen. This is not unexpected, because the chiral centre of ketorolac resides within an alicyclic ring, which would make the possibility of inversion unlikely. Since all pharmacokinetic indices, particularly the Vd, were similar for both enantiomers, it is likely that the 2 enantiomers are equally bound to plasma proteins. From this preliminary study, it might be concluded that the pharmacokinetic data derived from 100

o R-ketorolac • S-ketorolac

1.00

80 60

0.10

40

::J

l

5

~ ~ c ~

8

20

8

I

0.01

"0

c

0

'0

~ GI

100

0&(

1.00

80 60

0.10

40 20

b 0.01

d

-1---.--.---.-...---,--...--,----,

o

2

4 Time (h)

6

8

15

20

25

Time (h)

Fig. 6. Plasma concentration and cumulative urinary excretion (Ae) vs time curves for ketorolac enantiomers after intravenous and oral doses ofketorolac tromethamine IOmg. Panels a and c show results after intravenous administration; band d show those after oral administration.

Pharmacokinetics of Ketorolac

423

Table II. Pharmacokinetic of ketorolac enantiomers in a healthy 45y male volunteer given 10mg of ketorolac tromethamine as the racemate Route

Intravenous Oral

AUC (mg/L· h)

t'12,8 (h)

CL (L/h/kg)

F(%)

Vd (L/kg)

R

S

R

S

R

S

R

S

R

S

2.35 2.00

2.27 1.92

2.3 2.1

2.1 2.4

0.019 0.023

0.020 0.024

0.066 0.070

0.062 0.081

100 85

100 85

Abbreviations: see table I.

studies which have used nonstereoselective assays may be used to explain the pharmacokinetics of the active S enantiomer, if the administered dose and AUe are divided by a factor of 2. This assumes, of course, that little interindividual variability is present in the S: R ratio of AUe values amongst humans, as this report deals with only 1 individual. Also, we cannot exclude the possibility of enantioselective disease-induced changes in pharmacokinetics of the enantiomers. It is also of note that the pharmacokinetics of ketorolac enantiomers are not stereoselective in the rat (Jamali et al. 1989b). This may make the rat a suitable animal model for the study of the pharmacokinetics of the enantiomers of ketorolac. Other chiral NSAIDs which show little or no stereoselectivity in humans include ketoprofen (Jamali & Brocks 1990) and tiaprofenic acid (Singh et a1. 1986).

4. Implications of Pharmacokinetic Properties for Therapeutic Use 4.1 Dosage and Therapeutic Range The recommended dosage of ketorolac tromethamine is 10mg orally every 4 to 6h for pain, up to a maximum of 40mg daily. For intramuscular use in the treatment of pain associated with surgical procedures, the initial recommended dosage is 30mg, followed by 10 to 30mg every 4 to 6h as needed, up to a daily maximum of 120mg (Anon. 1991). The relationship between plasma concentrations ofketorolac and effect have not been studied. There is a potential for such a relationship, as there is some evidence for an analgesia-plasma concentration relationship for other NSAIDs (D' Arienzo

et a1. 1984; Kohler et a1. 1985; Laska et al. 1986). Nevertheless, by examining the time-effect curves for ketorolac obtained from other studies, some inferences can be made. For example, Forbes et a1. (1990) described a double-blind study involving 161 patients who had dental surgery for removal of at least 1 third molar. In a single dose experiment, each patient took a single postoperative dose of either placebo, aspirin, acetaminophen with codeine or 10mg of ketorolac tromethamine at the postoperative onset of moderate to severe pain. The patients self-monitored their pain status for 6h after taking the analgesic capsule. For patients receiving ketorolac, mean maximal pain relief was achieved after 3h (fig. 7). This is in contrast to the pharmacokinetic data of ketorolac in healthy volunteers, in which the time of maximal plasma concentration is reached in less than Ih (table I). Since ketorolac is a peripherally acting analgesic (Rooks et a1. 1982, 1985), the delay in analgesia relative to plasma concentrations may represent a relatively slow rate of distribution of the drug to the site of action, which in this case is the dental gums and surrounding tissues: The rate of transfer of ketorolac into the tissues may be slow due to its high degree of protein binding in plasma (section 2.2.1). It is known that the t max of NSAIDs in the synovial fluid lags behind that in plasma or serum (Netter et al. 1989). Alternatively, a delay in t max is also plausible in patients after dental surgery. In another study involving patients who had undergone major surgery (Stanski et a1. 1990) and received intramuscular ketorolac tromethamine 30 to 90mg, the time-effect curve for mean pain relief also indicated a somewhat delayed onset of pain relief (60 to 120 min) compared with the t max ob-

424

Clin. Pharmacokinet. 23 (6) 1992

3

2.2.1). The investigators did not state whether the differences between the 1 and 4h samples were different within age groups. Judging from the mean values, there may be a trend towards a greater unbound fraction in the earlier sampling times. This may be indicative of saturable binding at higher concentrations, as discussed in section 2.2.1. Unfortunately, the study by Jallad et al. (1990) did not examine urinary excretion of ketorolac, and clinical parameters such as creatinine clearance and albumin concentration in plasma were not reported . In a recent study (Olkkola & Maunuksela 1991), the pharmacokinetics of ketorolac were studied in children (age 6.0 ± 1.2 years) given intravenous ketorolac tromethamine 0.5 mg/kg postoperatively. The mean t'hil was similar to that observed in healthy adult volunteers. However, the mean Vd and CL were both higher. The investigators suggested that the larger Vd in children was due to either a higher body water volume and/or a lesser degree of plasma protein binding than in adults. The latter explanation could also explain the higher CL seen in the children. It was concluded that a higher dose of ketorolac may be required in children to compensate for the higher CL, although this may not be necessary if there was also an increase in the unbound fraction in plasma.

,,

I"~

".,.,................. ... Paracetamol 600mg + codeine 60mg (n = 27)

/

.///"sPirin 650mg " 1/ (n = 32)

, i ,,/ .....

.........

'"

.......................................

/ ......... Placebo (n = 32)

.:....

O~--~r----r----.----.-----.--~

o

2

3

5

4

6

Time (h)

Fig. 7. Mean pain relief vs time profiles after dental surgery in patients receiving oral doses of ketorolac tromethamine IOmg alone, paracetamol 600mg + codeine 60mg in combination, aspirin 650mg alone or placebo alone (from Forbes et al. 1990, with permission).

tained from pharmacokinetic studies volunteers (37 to 58 min).

10

healthy

4.2 Influence of Age on Ketorolac Pharmacokinetics Since the disposition of ketorolac is heavily dependant on intrinsic hepatic clearance and protein binding, conditions such as aging and renal or hepatic diseases might possibly influence the pharmacokinetics of the drug. The effects of aging on the ketorolac pharmacokinetics were explored by Jallad et al. (1990) in healthy volunteers given single doses ofketorolac tromethamine lOmg oral tablets or 30mg intramuscular injections. Both a young and an elderly group were examined and the pharmacokinetics were compared (table I). There were statistically significant increases in AUC and t max after oral doses and reduced CL after intramuscular doses in the elderly compared with the young volunteers. In the elderly, the plasma protein binding ofketorolac following intramuscular 30mg doses was 98.9 and 99.4% after 1 and 4h postdose, respectively. The lh samples, but not the 4h samples, were significantly different from those obtained in a matching group of young volunteers (see section

4.3 Effects of Renal and Hepatic Diseases Information regarding the influence of disease on the pharmacokinetics of ketorolac is limited to 2 published abstracts which describe the pharmacokinetics of ketorolac in the presence of renal or hepatic disease. Martinez et al. (1987) examined 10 patients with renal insufficiency and compared the pharmacokinetics of ketorolac with those in 16 young healthy volunteers (table I). The extent of the renal insufficiency in the patients was not reported. The presence of renal impairment resulted in significantly greater AUC and t'l2il values, and a complimentary reduction in CL of ketorolac, when either oral lOmg or intramuscular 30mg doses were given. The t max following oral tablets was also prolonged in the renal patients. There was no explanation as to how ketorolac accumulates in plasma,

Pharmacokinetics of Ketorolac

considering its extensive metabolism by the liver (Mroszczak et al. 1987). It appears that the investigators did not examine the possibility of accumulation of conjugated ketorolac in plasma or the urinary excretion of conjugated ketorolac. Although it is assumed that conjugated drug is inactive, it would represent a source of active drug in the event of hydrolysis of conjugates in vivo (Brocks et al. 1991; Verbeeck et al. 1984). Alternatively, the accumulation of conjugates might allow more drug to be excreted in the bile. This would provide a source of drug that could be reintroduced into the body via enterohepatic recirculation (Levy 1979). Nevertheless, the investigators suggested the need for reduced dosages and close monitoring of renal patients for their responses to ketorolac. The effects of hepatic disease on ketorolac pharmacokinetics were studied by Pages et al. (1987) in 7 male patients with hepatic impairment of undisclosed aetiology and severity (table I). Ketorolac tromethamine oral IOmg and intramuscular 30mg doses were given. There were no differences between the patients and a group of healthy young male volunteers in AUC, Cmax , or CL (table I). However, t max and t'12(:1 were significantly higher in the patients following the oral and intramuscular doses, respectively. Although the pharmacokinetic differences in the patients were slight, as with the study by Martint;z et al. (1987), the investigators suggested the need for close monitoring of patients with hepatic disease who receive ketorolac.

5. Pharmacokinetic Drug Interactions 5.1 Effect of Other Drugs on the Pharmacokinetics of Ketorolac There is limited information available regarding the effects of other drugs on the pharmacokinetics of ketorolac. The manufacturer has stated that therapeutic concentrations of digoxin, warfarin, paracetamol, phenytoin, tolbutamide and piroxicam have no effect on the plasma protein binding of ketorolac (Anon. 1991). However, these studies have not been published to date. To counteract the adverse gastrointestinal side effects of NSAIDs, antacids are sometimes coad-

425

ministered. In the Mroszczak et al. (1990) review, a study conducted by Syntex was described which examined the effect of a commonly used magnesium and aluminum hydroxide containing antacid ('Maalox', Rorer) on the pharmacokinetics of ketorolac (fig. 3). The antacid had no effect on the pharmacokinetics of a single 10mg dose of ketorolac tromethamine given to 12 healthy volunteers (6 men, 6 women). 5.2 Effect of Ketorolac on the Pharmacokinetics of Other Drugs The pharmacokinetics of a number of drugs are known to be changed with the coadministration of different NSAIDs. Such interactions have been seen between NSAIDs and medications including oral anticoagulants, oral hypoglycaemics, digoxin, anticonvulsants, lithium and methotrexate (Verbeeck 1990). Consequently, it might be recommended that therapy be closely monitored when ketorolac is given together with drugs known to interact with NSAIDs. There have, unfortunately, been very few published reports specifically involving the potential interactions between ketorolac and other medications. Warfarin may be used as a postoperative anticoagulant in patients and it is known that the S enantiomer of warfarin is 5 times more potent than its R antipode. Toon et al. (1990) studied the possibility of an interaction between ketorolac and the enantiomers of warfarin after single doses of warfarin administered as the racemate. 12 healthy young male volunteers (20 to 32 years) were given either placebo or ketorolac 10mg 4 times daily for 6 days in a double-blind 2-way randomised crossover design. On day 6, a dose of 25mg of racemic warfarin was given and warfarin concentrations and haematological changes were assessed. The investigators also studied the possible displacement of warfarin from plasma proteins, using ultracentrifugation, and 14C-Iabelled racemic warfarin added to peak and trough warfarin plasma samples. An assessment of the relative protein binding of the warfarin enantiomers was not performed using this method.

426

The administration ofketorolac caused the mean concentrations of both warfarin enantiomers to be consistently lower than placebo. AVC values were not reported and the calculated CL values, which were not normalised for bodyweight, did not differ between treatments. However, a significantly reduced Vd was noted for the R enantiomer of warfarin after administration of ketorolac. For S-warfarin, there was a significant reduction in the Cmax , but no change in the Yd. No effect ofketorolac was seen on the plasma protein binding of (R,S)-warfarin. With respect to the pharmacodynamic effects ofketorolac, it did not alter the haemostatic effects of warfarin. The investigators concluded that there was a minimal effect of ketorolac on the pharmacokinetics of warfarin (Toon et al. 1990). However, they suggested the need for careful monitoring of warfarin therapy because of their use of healthy volunteers rather than patients and because only single doses of warfarin were given. Another consideration that was not addressed by the investigators is that a relatively small dose of ketorolac was used in this study. Although multiple doses of 10mg were given, it must be considered that postoperatively ketorolac has been recommended at intramuscular doses up to 30mg. Therefore, immediately postoperatively the risks of a drug interaction between ketorolac and warfarin may be increased due to the higher concentrations of ketorolac incurred with higher doses.

6. Conclusions Ketorolac is an NSAID which is indicated for analgesia. It is especially beneficial due to its availability as a parenteral formulation, which may allow it to be used as a replacement for opiate analgesics postoperatively and during labour. Ketorolac displays many of the same pharmacokinetic characteristics as other NSAIDs. There has been little if any attempt reported to date to establish a relationship between plasma ketorolac concentrations and analgesia. This would be an important area in which to direct future research regarding ketorolac. In preparing this review article, it was evident

Clin. Pharmacokinet. 23 (6) 1992

that much of the information related to the pharmacokinetics of ketorolac have been derived from manufacturer's studies which have been publicised in the form of abstracts from scientific meetings; much of the data is confined to manufacturer's files. An example is the assay procedures which have been used by Syntex to quantify concentrations of ketorolac in biological samples; none of the assays have been published in the form of full papers. This is unfortunate and it is hoped that the manufacturer will publish more of their data regarding ketorolac in the future in peer-reviewed scientific journals.

References Ahn HY, Jamali F, Cox SR, Kittayanond D, Smith DE. Stereoselective disposition of ibuprofen enantiomers in the isolated perfused rat kidney. Pharmaceutical Research 8: 1520-1524, 1991 Anonymous. Toradol prescribing information. Syntex Inc. and Upjohn Co. of Canada. 1991 Brocks DR, Jamaii F, Russell AS. Stereoselective disposition of etodolac enantiomers in synovial fluid. Journal of Clinical Pharmacology 31: 741-746, 1991 Buckley MMT, Brogden RN. Ketorolac: a review of its pharmacodynamic and pharmacokinetic properties, and therapeutic potential. Drugs 39: 86-109, 1990 D'Arienzo M, Pennisi M, Zanolo G, Borsa M. Ketoprofen lysine: ketoprofen serum levels and analgesic activity. Drugs in Experimental Oinical Research 10: 863-866, 1984 Debruyne D, de Ligny H, Ryckelynck J, Albessard F, Moulin M. Clinical pharmacokinetics of ketoprofen after single intravenous administration as a bolus or infusion. Clinical Pharmacokinetics 12: 214-221, 1987 Flach AJ, Lavelle CJ, Olander KW, Retzlaff JA, Sorenson LW. The effect of ketorolac tromethamine solution 0.5% in reducing postoperative inflammation after cataract extraction and intraocular lens implantation. Ophthalmology 95:1279-1284, 1988 Forbes JA, Butterworth GA, Burchfield WH, Beaver AT. Evaluation ofketorolac, aspirin, and acetaminophen-codeine combination in postoperative oral surgery pain. Pharmacotherapy 10: 77S-93S, 1990 Guzman A, Yuste F, Toscano RA, Young JM, Van Hom AR. Absolute configuration of (-)o5-benzoyl-1 ,2-dihydro-3H-pyrrolo[I,2-a]pyrrole-l-carboxylic acid, the active enantiomer of ketorolac. Journal of Medicinal Chemistry 29: 589-591, 1986 Jallad NS, Garg DC, Martinez JJ, Mroszczak EJ, Weidler DJ. Pharmacokinetics of single-dose oral and intramuscular ketorolac tromethamine in the young and elderly. Journal ofOinical Pharmacology 30: 76-81, 1990 Jamaii F, Stevens DRS. Naproxen excretion in milk and its uptake by the infant. Drug Intelligence and Clinical Pharmacy 17: 910-911, 1983 Jamali F. Pharmacokinetics of enantiomers of chiral non-steroidal anti-inflammatory drugs. European Journal of Drug Metabolism and Pharmacokinetics 13: 1-9, 1988 Jamali F, Mehvar R, Pasutto FM. Enantioselective aspects of drug action and disposition: therapeutic pitfalls. Journal of Pharmaceutical Sciences 78: 695-715, 1989a

Pharmacokinetics of Ketorolac

Jamali F, Pasutto FM, Lemko C. HPLC ofketorolac enantiomers and application to pharmacokinetics in the rat. Journal of Liquid Chromatography 12: 1835-1850, 1989b Jamali F, Brocks DR. Clinical pharmacokinetics of ketoprofen and its enantiomers. Clinical Pharmacokinetics 19: 197217,1990 Jung D, Mroszczak EJ, Bynum L. Pharmacokinetics ofketorolac tromethamine in humans after intravenous, intramuscular, and oral administration. European Journal of Clinical Pharmacology 35: 423-425, 1988 Jung D, Mroszczak EJ, Wu AT, Ling TL, Sevelius H, et a!. Pharmacokinetics of ketorolac and p-hydroxyketorolac following oral and intramuscular administration of ketorolac tromethamine. Pharmaceutical Research 6: 62-65, 1989 Kanto J. Obstetric analgesia: clinical pharmacokinetic considerations. Clinical Pharmacokinetics II: 283-298, 1986 Kohler G, Primbs P, Morand J, Rubelt CH. Correlation between ketoprofen plasma levels and analgesic effect in acute lumbar pain and radicular pain. Clinical Rheumatology 4: 399-404, 1985 Laska EM, Sunshine A, Marrero I, Olson N, Siegel C, et a!. The correlation between blood levels of ibuprofen and clinical analgesic response. Clinical Pharmacology and Therapeutics 40: 17, 1986 Levy G. Decreased body clearance of diflunisal in renal insufficiency - an alternate explanation. British Journal of Clinical Pharmacology 8: 601, 1979 Lin JH, Cocchetto DM, Duggan DE. Protein binding as a primary determinant of the clinical pharmacokinetic properties of nonsteroidal anti-inflammatory drugs. Clinical Pharmacokinetics 12: 402-432, 1987 Ling TL, Combs DL. Ocular bioavailability and tissue distribution of [14C]ketorolac tromethamine in rabbits. Journal of Pharmaceutical Sciences 76: 289-294, 1987 Litvak KM, McEvoy GK. Ketorolac, an injectable non-narcotic analgesic. Clinical Pharmacy 9: 921-935, 1990 Martinez J, Garg DC, Pages U, Jallad NS, Vee JP, et a!. Single dose pharmacokinetics of ketorolac in healthy young and renal impaired subjects. Abstract No. 72. Journal of Clinical Pharmacology 27: 722, 1987 Mroszczak EJ, Ling T, Vee J, Massey I, Sevelius H. Ketorolac tromethamine absorption and pharmacokinetics in humans. Clinical Pharmacology and Therapeutics 37: 215, 1985 Mroszczak EJ, Lee FW, Combs D, Sarnquist FH, Huang B, et a!. Ketorolac tromethamine absorption, distribution, metabolism, excretion, and pharmacokinetics in animals and humans. Drug Metabolism and Disposition 15: 618-626, 1987 Mroszczak EJ, Jung D, Vee J, Bynum L, Sevelius H, et a!. Ketorolac tromethamine pharmacokinetics and metabolism after intravenous, intramuscular, and oral administration in humans and animals. Pharmacotherapy 10: 33S-39S, 1990 Netter P, Bannwarth B, Royer-Morrot MJ. Recent findings on the pharmacokinetics of non-Steroidal anti-inflammatory drugs in synovial fluid. Clinical Pharmacokinetics 17: 145-162, 1989 Notarianni U, Oldham HG, Bennett PN. Passage ofparacetamol into breast milk and its subsequent metabolism by the neonate. British Journal of Clinical Pharmacology 24: 63-67, 1987 Olkkola KT, Maunuksela E. The pharmacokinetics of postoperative intravenous ketorolac tromethamine in children. British Journal of Clinical Pharmacology 31: 182-184, 1991 Pages U, Martinez JJ, Garg DC, Vee JP, Mroszczak J, et a!. Pharmacokinetics of ketorolac tromethamine in hepatically impaired vs young healthy subjects. Abstract No. 78. Journal of Clinical Pharmacology 27: 724, 1987 Resman-Targoff BH. Ketorolac: a parenteral nonsteroidal antiinflammatory drug. Drug Intelligence and Clinical Pharmacy 24: 1098-1104, 1990 Reynolds J. (Ed.) Analgesic and antiinflammatory drugs. In: Mar-

427

tindale the extra pharmacopoeia, 29th ed., pp. 1-46, Pharmaceutical Press, London, 1989 Rooks WH II, Tomolonis AJ, Maloney PJ, Wallach MB, Schuler ME. The analgesic and anti-inflammatory profile of (+ )-5-benzoyl-I,2-