Journal of Analytical Toxicology, Vol. 14, JuiylAugust 1990
PercutaneousAbsorption of TricyclicAntidepressants: Amitriptyline, Nortriptyline, imipramine, and Desipramine D a v i d N. B a i l e y *
Division of Laboratory Medicine, Department of Pathology, University of California Medical Center, San Diego, California 92103
Abstract I The percutaneous absorption of amitriptyline, nortrlptyline, imipramine, and desipramlne as their hydrochlorlde salts In vivo was demonstrated without use of s vehicle using the hairless (hr-1/hr-1) mouse as an experimental model for human skin. After topical application of 2 mg of each compound in distilled water, followed by rapid evaporation of the water, concentrations were measured in heart, lung, brain, liver, and blood in 1-, 2-, 4-, and 6-hour study groups9 Lung consistently demonstrated the highest concentrations for all four compounds while heart and liver had the lowest. Concentrations in heart remained essentially constant for all compounds during the 6-hour study period. The concentrations in solid tissues were much lower than those commonly seen In man after overdose, whereas the concentrations in blood resembled low therapeutic to toxic concentrations in humans. Percutaneous absorption may provide s feasible route of administration for the trlcyclic antidepressants which may lead to Improved compliance with fewer gastrointestinal side effects.
Introduction The transdcrmal route of administration has been investigated for a wide variety of drugs including steroids such as hydrocortisone (1) and testosterone (2), antihypertensive agents such as minoxidil (3), and narcotic analgesics such as meperidine (4) and methadone (5). In addition, some compounds such as scopolamine and trinitroglycerin have been commercially available for many years as transdermal preparations (6). Since this route requires relatively little motivation on the part of the patient, ensures essentially constant administration of the agent, and bypasses the gastrointestinal tract as a site of absorption, it might provide a useful means of administering drugs to those patients for whom compliance may be a problem, for whom continuous drug administration is desirable, and for whom gastrointestinal side effects may occur. One such group of patients is that for whom the tricyclic antidepressants are prescribed. With this group in 9Address correspondence and reprint requests to David N. Bailey. M.D., Division of Laboratory Medicine, Department of Pathology, UCSD Medical Center (H.720-T), 225 Dickinson Street, San Diego, California 92103.
mind, the percutaneous absorption of amitriptyline and imipramine along with their N-desmethylated analogs, nortriptyline and desipramine, was investigated using the hairless (hr-1/hr-1) mouse as an experimental model for human skin (7,8). The hydrochloride salts rather than the free bases were studied because of their excellent aqueous solubility and relative stability.
Experimental
Animals. Male, hairless (hr-l/hr-l) mice, four to six weeks old, were purchased from Simonsen Laboratories. Dosing solutions. Solutions of amitriptyline, nortriptyline, imipramine, and desipramine as their hydrochlorides were each prepared in distilled water at 100 mg/mL for amitriptyline and imipramine and at 50 mg/mL for nortriptyline and dcsipramine. Protocol. For each compound and for each study hour (1, 2, 4, and 6 hours) a volume of dosing solution sufficient to yield 2 mg of drug (either 20 or 40 #L) was applied topically to a l-cm 2area (107oof total body surface area of 70 cm 2) (9) of the back of each of 10 mice which had been immobilized atraumatically in such fashion that they could not access the dosing site. The solution was dried within one minute, leaving a thin residue of drug which was maintained free from touch. For each compound and for each study hour, 10 additional mice were treated topically with equal volumes of distilled water. At the conclusion of each timed study period, mice were sacrificed with carbon dioxide asphyxiation. Heart, lung, brain, and liver were excised in toto, and each respective organ was pooled, weighed, and homogenized with 6 mL of water per gram of tissue wet weight. Each homogenate was centrifuged at 3,000 g for 30 min, and the supernatant fluid was harvested. After decapitation of each mouse, blood was also pooled into a tube containing potassium oxalate and sodium fluoride. The tissue homogenates and whole blood were frozen ( - 20~ until the time of analysis, which was within 7 to 10 days of each experiment. Drug analysis. Tissue homogenates and whole blood were extracted at pH 10 into hexane-isoamyl alcohol (98:2, v/v) conraining ioxapine as an internal standard and were back-extracted into 0.IN HCI. After alkalinization of the acid extract, each tricyclic antidepressant was reextracted into hexane-isoamyl alcohol, and the extract was evaporated. Following addition of mobile phase (10-mmol/L phosphate buffer-acetonitrile, 3:1 v/v
Reproduction (photocopying) of editonol content o! this journal is prohibited without publisher's permission.
217
Journal of AnalyticalToxicology,Vol. 14, July/August 1990
for amitriptyline and nortriptyline and 3.8:1 v/v for imipramine and desipramine), each residue was submitted to isocratic, reversed-phase high-pressure liquid chromatography with ultraviolet detection (200 nm). Drug concentrations were calculated using peak-height ratios (drug-internal standard) relative to those of reference calibrators (prepared by in vitro supplementation of pooled drug-free mouse tissue homogenates and blood obtained from the water-dosed mice). Only the respective unchanged drugs were measured. Drug concentrations in each tissue were calculated from the concentration in the homogenate, the total volume of homogenate prepared, and the mass of the excised tissue utilized. Because the respective tissues had been pooled (due to the extremely small mass of each individual tissue and the anticipated low drug concentrations), these values represent mean tissue concentrations. The analytical sensitivity (limit of detection) of the procedure is 30 ng/g of solid tissue wet weight and 5 ng/mL of blood.
Results and Discussion The percutaneous absorption of amitriptyline (MW 277), nortriptyline (MW 263), imipramine (MW 280), and desipramine (MW 266) as their hydrochloride salts has been demonstrated in vivo without use of a vehicle. Chromatographic peaks corresponding to unchanged drug were detected only in the study material and were not detected in mice treated with water. The tissue distribution of each unchanged drug is shown in Figure I. The highest concentrations over time were consistently found in lung (ranging from 931 to 9,950 nmol/Kg) (approximately 250-2,700 ng/g, using an average molecular weight of 272 for the tricyclic antidepressants), and the lowest were detected in heart (ranging from "not detected" to 606 nmol/Kg) ("not detected" to approximately 165 ng/g) and liver (ranging from "not detected" to 1,300 nmol/Kg) ("not detected to approximately 350 ng/g). Concentrations in brain (ranging from 140 to 8,240 nmol/Kg) (approximately 40-2,250 ng/g) and blood (ranging from 89 to 3,810 nmol/L) (approximately 25-I,050
10,000 7,500 5,000 2,500
O
10,000~
C 7,500 5~00 2,500
1
2
3
4
S
6
Hours
Figure 1. Mean tissue concentrations following topical applicationof 2 mg of tricyclicantidepressantsas theirhydrochloridesalts:(A) amitripP/line, (8) nortriptyline,(C) imiprarnine,and (0) desipramine.Key;(0) heart, (O) lung, (11)brain, ([3) liver, and (A) blood.All concentrations are expressedin nmol/Kg tissuewet weightexceptfor blood (nmol/L).
218
ng/mL) were approximately intermediate. Concentrations in heart tended to remain constant for all four tricyclic antidepressants over time. The concentrations in solid tissues were much lower than those commonly reported after overdose in humans (I0), whereas the concentrations in blood resembled low therapeutic to toxic concentrations in humans. Conclusion
Percutaneous absorption of tricyclic antidepressants may be a feasible route of administration of these drugs. While the compounds as their more soluble and stable hydro,chloride salts were applied in aqueous solution, the water was evaporated within one minute so that the transdermal absorption effectively occurred without use of a vehicle. Such absorption could conceivably even be enhanced and better controlled by use of the unionized (free base) drug solubilized in an appropriate vehicle. This route may improve compliance and maintain continuous administration of drug. Although the most common side effects of tricyclic antidepressant therapy (sedation, hypotension, and anticholinergic effects) (6) would likely not be ameliorated by transdermal administration, this route would probably reduce the incidence of such side effects as nausea, diarrhea, and stomatitis (6), which have been associated with oral ingestion. These factors may be especially important given the population typically treated with these compounds. Studies in human volunteers should be undertaken to validate these hypotheses.
Acknowledgment
The author gratefully acknowledges the expert technical assistance of Joseph J. Coffee in performing the experiments described in this manuscript.
References 1. R.C. Wester, P.K. Noonan, and H.I. Maibach. Frequency of application on percutaneous absorption of hydrocortisone. Arch. Dermatol. 113:620-22 (1977). 2. R.C. Waster, RK. Noonan, and H.I. Maibach. Variations in percutaneous absorption of testosterone in rhesus monkey due to anatomic site of application and frequency of application. Arch. Dermatol. Res. 267:229-35 (1980). 3. M.G. Eller, G.J. Szpunar, and A.A. Della-Coletta. Absorption of minoxidil after topical application: effect of frequency and site of application. Clin. Pharmacol. Thor. 45:396.402 (1989). 4. D.N. Bailey, J.J. Coffee, and J.R. Briggs. Tissue distribution of meperidine following percutaneous absorption as its hydrochlofide salt in vivo. J. Toxicol.-Cut. & Ocular ToxicoL 6:227-32 (1987). 5. D.N. Bailey, J.R. Briggs, and J.J. Coffee. Tissue distribution of methadone following percutaneous absorption in vivo. J. Toxi. col.-Cut. & Ocular Toxicol. 5:303-308 (1986). & Physicians' Desk Reference. Medical Economics, Oradell, NJ, 1989, pp. 864, 866. 7. H,C. Brooke. Hairless mica. J. Herod. 17:173-74 (1926). & M.C. Green. In The Biology of the Laboratory Mouse, E.L. Green, Ed., McGraw Hill, New York, 1966, p. 87. 9. J.K. Inglis. Introduction to Laboratory Animal Science and Technology. Pergamon, New York, 1980, p. 255. 10. R.C. Basalt and R.H. Cravey. Disposition of Toxic Drugs and Chemicals in Man. Year Book Medical Publishers, Chicago, 1989, pp. 38.43, 423-26. Manuscript received November 13, 1989; revision received January 11, !990.
PercutaneousAbsorption of TricyclicAntidepressants: Amitriptyline, Nortriptyline, imipramine, and Desipramine D a v i d N. B a i l e y *
Division of Laboratory Medicine, Department of Pathology, University of California Medical Center, San Diego, California 92103
Abstract I The percutaneous absorption of amitriptyline, nortrlptyline, imipramine, and desipramlne as their hydrochlorlde salts In vivo was demonstrated without use of s vehicle using the hairless (hr-1/hr-1) mouse as an experimental model for human skin. After topical application of 2 mg of each compound in distilled water, followed by rapid evaporation of the water, concentrations were measured in heart, lung, brain, liver, and blood in 1-, 2-, 4-, and 6-hour study groups9 Lung consistently demonstrated the highest concentrations for all four compounds while heart and liver had the lowest. Concentrations in heart remained essentially constant for all compounds during the 6-hour study period. The concentrations in solid tissues were much lower than those commonly seen In man after overdose, whereas the concentrations in blood resembled low therapeutic to toxic concentrations in humans. Percutaneous absorption may provide s feasible route of administration for the trlcyclic antidepressants which may lead to Improved compliance with fewer gastrointestinal side effects.
Introduction The transdcrmal route of administration has been investigated for a wide variety of drugs including steroids such as hydrocortisone (1) and testosterone (2), antihypertensive agents such as minoxidil (3), and narcotic analgesics such as meperidine (4) and methadone (5). In addition, some compounds such as scopolamine and trinitroglycerin have been commercially available for many years as transdermal preparations (6). Since this route requires relatively little motivation on the part of the patient, ensures essentially constant administration of the agent, and bypasses the gastrointestinal tract as a site of absorption, it might provide a useful means of administering drugs to those patients for whom compliance may be a problem, for whom continuous drug administration is desirable, and for whom gastrointestinal side effects may occur. One such group of patients is that for whom the tricyclic antidepressants are prescribed. With this group in 9Address correspondence and reprint requests to David N. Bailey. M.D., Division of Laboratory Medicine, Department of Pathology, UCSD Medical Center (H.720-T), 225 Dickinson Street, San Diego, California 92103.
mind, the percutaneous absorption of amitriptyline and imipramine along with their N-desmethylated analogs, nortriptyline and desipramine, was investigated using the hairless (hr-1/hr-1) mouse as an experimental model for human skin (7,8). The hydrochloride salts rather than the free bases were studied because of their excellent aqueous solubility and relative stability.
Experimental
Animals. Male, hairless (hr-l/hr-l) mice, four to six weeks old, were purchased from Simonsen Laboratories. Dosing solutions. Solutions of amitriptyline, nortriptyline, imipramine, and desipramine as their hydrochlorides were each prepared in distilled water at 100 mg/mL for amitriptyline and imipramine and at 50 mg/mL for nortriptyline and dcsipramine. Protocol. For each compound and for each study hour (1, 2, 4, and 6 hours) a volume of dosing solution sufficient to yield 2 mg of drug (either 20 or 40 #L) was applied topically to a l-cm 2area (107oof total body surface area of 70 cm 2) (9) of the back of each of 10 mice which had been immobilized atraumatically in such fashion that they could not access the dosing site. The solution was dried within one minute, leaving a thin residue of drug which was maintained free from touch. For each compound and for each study hour, 10 additional mice were treated topically with equal volumes of distilled water. At the conclusion of each timed study period, mice were sacrificed with carbon dioxide asphyxiation. Heart, lung, brain, and liver were excised in toto, and each respective organ was pooled, weighed, and homogenized with 6 mL of water per gram of tissue wet weight. Each homogenate was centrifuged at 3,000 g for 30 min, and the supernatant fluid was harvested. After decapitation of each mouse, blood was also pooled into a tube containing potassium oxalate and sodium fluoride. The tissue homogenates and whole blood were frozen ( - 20~ until the time of analysis, which was within 7 to 10 days of each experiment. Drug analysis. Tissue homogenates and whole blood were extracted at pH 10 into hexane-isoamyl alcohol (98:2, v/v) conraining ioxapine as an internal standard and were back-extracted into 0.IN HCI. After alkalinization of the acid extract, each tricyclic antidepressant was reextracted into hexane-isoamyl alcohol, and the extract was evaporated. Following addition of mobile phase (10-mmol/L phosphate buffer-acetonitrile, 3:1 v/v
Reproduction (photocopying) of editonol content o! this journal is prohibited without publisher's permission.
217
Journal of AnalyticalToxicology,Vol. 14, July/August 1990
for amitriptyline and nortriptyline and 3.8:1 v/v for imipramine and desipramine), each residue was submitted to isocratic, reversed-phase high-pressure liquid chromatography with ultraviolet detection (200 nm). Drug concentrations were calculated using peak-height ratios (drug-internal standard) relative to those of reference calibrators (prepared by in vitro supplementation of pooled drug-free mouse tissue homogenates and blood obtained from the water-dosed mice). Only the respective unchanged drugs were measured. Drug concentrations in each tissue were calculated from the concentration in the homogenate, the total volume of homogenate prepared, and the mass of the excised tissue utilized. Because the respective tissues had been pooled (due to the extremely small mass of each individual tissue and the anticipated low drug concentrations), these values represent mean tissue concentrations. The analytical sensitivity (limit of detection) of the procedure is 30 ng/g of solid tissue wet weight and 5 ng/mL of blood.
Results and Discussion The percutaneous absorption of amitriptyline (MW 277), nortriptyline (MW 263), imipramine (MW 280), and desipramine (MW 266) as their hydrochloride salts has been demonstrated in vivo without use of a vehicle. Chromatographic peaks corresponding to unchanged drug were detected only in the study material and were not detected in mice treated with water. The tissue distribution of each unchanged drug is shown in Figure I. The highest concentrations over time were consistently found in lung (ranging from 931 to 9,950 nmol/Kg) (approximately 250-2,700 ng/g, using an average molecular weight of 272 for the tricyclic antidepressants), and the lowest were detected in heart (ranging from "not detected" to 606 nmol/Kg) ("not detected" to approximately 165 ng/g) and liver (ranging from "not detected" to 1,300 nmol/Kg) ("not detected to approximately 350 ng/g). Concentrations in brain (ranging from 140 to 8,240 nmol/Kg) (approximately 40-2,250 ng/g) and blood (ranging from 89 to 3,810 nmol/L) (approximately 25-I,050
10,000 7,500 5,000 2,500
O
10,000~
C 7,500 5~00 2,500
1
2
3
4
S
6
Hours
Figure 1. Mean tissue concentrations following topical applicationof 2 mg of tricyclicantidepressantsas theirhydrochloridesalts:(A) amitripP/line, (8) nortriptyline,(C) imiprarnine,and (0) desipramine.Key;(0) heart, (O) lung, (11)brain, ([3) liver, and (A) blood.All concentrations are expressedin nmol/Kg tissuewet weightexceptfor blood (nmol/L).
218
ng/mL) were approximately intermediate. Concentrations in heart tended to remain constant for all four tricyclic antidepressants over time. The concentrations in solid tissues were much lower than those commonly reported after overdose in humans (I0), whereas the concentrations in blood resembled low therapeutic to toxic concentrations in humans. Conclusion
Percutaneous absorption of tricyclic antidepressants may be a feasible route of administration of these drugs. While the compounds as their more soluble and stable hydro,chloride salts were applied in aqueous solution, the water was evaporated within one minute so that the transdermal absorption effectively occurred without use of a vehicle. Such absorption could conceivably even be enhanced and better controlled by use of the unionized (free base) drug solubilized in an appropriate vehicle. This route may improve compliance and maintain continuous administration of drug. Although the most common side effects of tricyclic antidepressant therapy (sedation, hypotension, and anticholinergic effects) (6) would likely not be ameliorated by transdermal administration, this route would probably reduce the incidence of such side effects as nausea, diarrhea, and stomatitis (6), which have been associated with oral ingestion. These factors may be especially important given the population typically treated with these compounds. Studies in human volunteers should be undertaken to validate these hypotheses.
Acknowledgment
The author gratefully acknowledges the expert technical assistance of Joseph J. Coffee in performing the experiments described in this manuscript.
References 1. R.C. Wester, P.K. Noonan, and H.I. Maibach. Frequency of application on percutaneous absorption of hydrocortisone. Arch. Dermatol. 113:620-22 (1977). 2. R.C. Waster, RK. Noonan, and H.I. Maibach. Variations in percutaneous absorption of testosterone in rhesus monkey due to anatomic site of application and frequency of application. Arch. Dermatol. Res. 267:229-35 (1980). 3. M.G. Eller, G.J. Szpunar, and A.A. Della-Coletta. Absorption of minoxidil after topical application: effect of frequency and site of application. Clin. Pharmacol. Thor. 45:396.402 (1989). 4. D.N. Bailey, J.J. Coffee, and J.R. Briggs. Tissue distribution of meperidine following percutaneous absorption as its hydrochlofide salt in vivo. J. Toxicol.-Cut. & Ocular ToxicoL 6:227-32 (1987). 5. D.N. Bailey, J.R. Briggs, and J.J. Coffee. Tissue distribution of methadone following percutaneous absorption in vivo. J. Toxi. col.-Cut. & Ocular Toxicol. 5:303-308 (1986). & Physicians' Desk Reference. Medical Economics, Oradell, NJ, 1989, pp. 864, 866. 7. H,C. Brooke. Hairless mica. J. Herod. 17:173-74 (1926). & M.C. Green. In The Biology of the Laboratory Mouse, E.L. Green, Ed., McGraw Hill, New York, 1966, p. 87. 9. J.K. Inglis. Introduction to Laboratory Animal Science and Technology. Pergamon, New York, 1980, p. 255. 10. R.C. Basalt and R.H. Cravey. Disposition of Toxic Drugs and Chemicals in Man. Year Book Medical Publishers, Chicago, 1989, pp. 38.43, 423-26. Manuscript received November 13, 1989; revision received January 11, !990.
