Am
J Psychiatry
BRIEF
135:3, March
COMMUNICATIONS
Prediction BY
THOMAS
1978
of Individual B. COOPER,
M.A.,
Dosage AND
GEORGE
of Nortriptyline M.
SIMPSON,
The authors describe a technique that enables a physician to determine individual patient dosage requirements for nortriptylinefrom a single 24-hour blood sample. Because the technique reveals immediately those patients at the extremes of dosage ranges, toxicity and the needfor time-consuming titration ofthe dosage regimen can be avoided.
IN 1971 WE REPORTED that individual dosage requirements could be predicted from a single dose of lithium carbonate by analyzing blood samples collected at 8 discrete time points during the 24 hours following ingestion of the drug (1). In 1973 we reported a major modification of this procedure that enabled us to predict individual dosage requirements from a single blood sample collected 24 hours after a 600-mg oral dose oflithium carbonate (2). The application and usefulness of that procedure were described in a 2-year follow-up study at this center (3). Seifert and associates (4) used our basic concept in determining a suitable regression equation for a slow-release formulation, and Chang and associates (5) recently confirmed our original observation using the 25-hour rather than the 24-hour time point. The debate over whether there is a therapeutic threshold (lower limit below which patients do not improve) or therapeutic window (lower threshold below which and upper limit above which patients do not improve) for the plasma concentration of the various antidepressant and neuroleptic drugs is far from resolution (6). Of all the data published to date the most consistent and convincing, in our opinion, are those
M.B.,
CH.B.
for nortriptyline. In endogenous depression only, a therapeutic window of 50-150 ng/ml has been claimed and confirmed (7-9). Asbemg (10) has recently shown that in patients randomly assigned to either a 25-mg t.i.d. or 50-mg t.i.d. regimen of nomtriptyline, the majority of their steadystate plasma levels will be within the therapeutic range on a dosage of 50 mg t.i.d. ; however, many patients who receive 25 mg t.i.d. will be below the lower plasma threshold of 50 ng/ml. Asberg recommended that patients for whom blood level determinations are not available should be given 50 mg t.i.d. and those who do not respond should have their medication decreased rather than increased. Alexandemson (1 1) has shown the ability to predict steady-state’ plasma levels of nomtriptyline by determining the plasma clearance rate from a single oral dose ofnomtmiptyline. The correlation between this mate and the steady-state plasma level was r=.93. These data clearly indicate that this system is linear. Additional evidence for linearity is the very high correlation (r= .99) between the plasma clearance mate calculated for the single oral dose study and the same rates calculated from the steady-state data. This linearity in the system suggests that the technique we have used for lithium prediction would be applicable to the prediction of the steady-state plasma levels of nomtriptyline. In addition, the constancy of this prediction over time was clearly demonstrated when the prediction obtained from a single-dose curve was applied 2 years later to the same patients who were then given nortriptyline and had steady-state values determined (11).
EXPERIMENTAL Mr. Cooper is Chief Investigator, Clinical Psychopharmacology Laboratory, Rockland Research Institute, Orangebung, N.Y. 10962, where Dr. Simpson was Principal Research Psychiatrist. Dr. Simpson is now Professor of Psychiatry and Director, U.S.C.-Metropsychopharmacology Services, Norwalk, Calif. This work was supported tal Health Administration tute of Mental Health.
in part by Alcohol, grant MH-08240
The unteers
0002-953X/78/0003-0333$0.40
©
subjects were 18 physically healthy (10 women and 8 men) aged 21-51
Steady state is defined been receiving a fixed
Drug Abuse, and Menfrom the National Insti-
DESIGN
American
Psychiatric
volwho
operationally as the time when a patient has dosage of medication such that repeat blood
level determinations carried do not differ significantly.
1978
normal years,
out
Association
each
day
before
the
morning
dose
333
BRIEF
Am
COMMUNICATIONS
fully understood the project and gave informed consent. Subjects fasted for 12 hours and then ingested 50 mg of nortmiptyline by mouth. Blood samples were collected immediately before ingestion of the medication (t0) and at 1,2,3,4,6,8,12,24, and 36 hours after this dose. The volunteers were not allowed to lie down for the first hour after ingestion and were not allowed to eat until a light meal was served 4 hours after ingestion of the medication. After the initial single-dose expemiments, all subjects ingested nortmiptyline on a 25-mg b.i.d. (9:00 a.m. and 9:00 p.m.) regimen for a period of 15 days, and blood samples were collected on days 7,14,15, and 16 (immediately before the morning dose of medication). The samples were analyzed in duplicate for nortriptyline content using a gas-liquid chromatography (GLC) procedure and a nitrogen detector (12). This method has a coefficient of variation less than 5% at a level of5 ng/ml and can detect less than I ng/ml when a 3-mI sample is extracted.
AND
DISCUSSION
/35:3,
March
1978
TABLE 1 Correlation Between Individual Time Points from Single-Dose and Steady-State Plasma Levels* Hours
After
50-mg
Dose
Correlation
-.002 +194
3
+316
4 6
+.282
1
8
+535 +555
12 24 36
+.839 +918 +963
*All
subjects
received
25 mg
b.i.d.
Study
Coefficient
2
(r)
of nortriptyline.
FIGURE 1 Plot of 24-Hour Data Versus Steady-State Plasma Level of 18 Volunteers After Ingestion of 25 mg of Nortriptyline b.i.d.* 140-.
y= l6.6+4.3(x)
.E110-J LU > LU -J
RESULTS
J Psychiatry
-
90-
Ui
z
80-
-J
The steady-state plasma level of nortriptyline determined on days 13,14, and 15 were within 10% of the average value for each individual patient. With one exception all subjects had attained the steady-state condition by the seventh day of continuous medication. Alexanderson ( I 1 ) made a similar observation 6 days after treatment began. Correlations between the mean steady-state plasma level and individual time points of the initial singledose curve are shown in table 1 The strong correlations at 12, 24, and 36 hours make it possible to use any one of these time points for accurate prediction of the steady-state level. For convenience we have chosen to use the 24-hour time point (figure 1); i.e., the patient returns for the blood collection at the same time on the next day. These data, reduced to a simple table of prediction of dosage level, are shown in table 2. Inspection of table 2 shows drug dosage changes detemmined by a 1 ng/ml blood level difference. A method with a coefficient of variation of 5% at best can be expected to measure a true level of 20 ng/ml as anything from 18-22 ng/ml. The change of dosage at, for example, 19 ng/ml is calculated to take into account this variance. Thus a true value of 19 ng/ml determined as 21 ng/ml will result in levels within the therapeutic range whether a 50-mg b.i.d. or 25-mg t.i.d. regimen is followed. Analysis of these samples in duplicate will reduce this variance further and is recommended as a routine procedure. If one accepts that a therapeutic window exists for nortriptyline, then by using this technique one can predict within 48 hours (assuming 24-hour turnaround in the laboratory) the precise dosage requirements for each patient. Thus a patient is given 50-mg of nortriptyline by the physician and instructed to return to the physician’s office (or a laboratory) to have a blood
700
60x
0
-
z 40F-
xx
2010(-I)
.
334
5
10
24-HOUR
PLASMA
0 *r_
15
20
25
NORTRIPTYLINE
LEVEL
.92.
TABLE
2
Dosage Regimen Based on 24-Hour gestion of 50 mg of Nortriptyline Suggested
Blood
Level
(mg/mI)
Blood Levels After In-
Suggested
Dosage
Regimen
41* *The
30 (ng/ml)
mg
10 mg ranges
calculated
sample laboratory instructed (during a blood can then increased ed when be kept
are for
calculated
these
by extrapolation
experimental
data
ofthe (figure
regression
b.i.d.
bid. line
(y=a+bx)
1).
drawn at the same time the next day. If the is not open at that time, the patient can be to take the 50-mg dose at some other time the working hours of the laboratory) and have sample collected 24 hours later. The patient be started on medication immediately (e.g., to 150 mg/day) and the dosage can be adjustlaboratory data are available. The patient can on this adjusted regimen for a suitable period
Am
J Psychiatry
135:3,
March
BRIEF
1978
(4 weeks), and if therapeutic response is not achieved, theme would appear to be little point in changing an already optimum dosage; rather, a change in medication or treatment modality must be considered by the clinician.
COMMENT
REFERENCES I.
Bergner P-EE, Berniker K, Cooper TB, Ct al: Lithium kinetics in man: effect of variation in dosage pattern. Br J Pharmacol 49:328-339, 1973 2. Cooper TB, Bergnen P-EE, Simpson GM: The 24-hour serum lithium level as a prognosticator of dosage requirements. Am J Psychiatry 130:601-603, 1973 3. Cooper TB, Simpson GM: The 24-hour lithium level as a prognosticator of dosage requirements: a 2-year follow-up study.
Am J Psychiatry
We emphasize that this observation applies only to a serious endogenous depression. A relationship between drug plasma level and clinical efficacy in nonendogenous depression has not been established (6). Despite the paucity of good clear data on the efficacy of monitoring plasma levels of drugs, the recent spate of publications of analytical procedures for tncyclic drugs from clinical laboratories seems to indicate that major efforts are under way to establish laboratories for routine screening using these measures. Given this situation, it would seem that the technique we have described may be of considerable help in a practical clinical setting, i.e., bringing a patient into therapeutic mange rapidly.2 The physician does not have to titrate the patient with medication, and the occasional patients who metabolize a drug extremely slowly or rapidly, and who therefore require very different drug regimens from the average patient, can be identified immediately. These combined elements may result in precise tailoring of the medication to each individual patient’s needs, thus minimizing overmedication that could result in untoward side effects. 2
Preliminary study ranging
of
data this from
from
technique 95 to 128
four have nglml
patients
participating
in a prospective
achieved steady state on the dosage predicted.
blood
levels
COMMUNICATIONS
133:440-443,
1976
4.
Seifert R, Bremkamp H, Junge C: Vereinfachte lithiumeinstellung dench belastungstest. Psychopharmacologia 43:285-286, 1975 5. Chang 55, Pandey GN, Casper R, et al: Predicting the optimal lithium dosage. Proceedings of the American Society for Clinical Pharmacology and Therapeutics. Dallas, Tex, March 23-25, 1977 6. Cooper TB, Simpson GM, Lee JH: Thymoleptic and Neuroleptic Drug Plasma Levels in Psychiatry: Current Status in International Review ofNeurobiology. vol 19. New York, Academic Press, 1976, pp 269-309 7. Asbeng M. Cronholm B, Sjoqvist F. et al: Relationship between plasma level and therapeutic effect of nortniptyline. Br Med J
3:331-334, 1971 8. Knagh-Sonensen tniptyline levels 9.
10.
P. Asbeng M, Eggert-Hansen in endogenous depression.
1973 Ziegler yE. Clayton levels and therapeutic 463, 1976 Asberg M: Treatment macokinetic and
kopsychiatry
9:18-26,
PJ.
Taylor response.
of depression pharmacodynamic
C: Plasma norLancet 1:113-115,
JR. et al: Nortriptyline Clin Pharmacol Then with
tnicyclic aspects.
plasma 20:458-
drugs-phanPhanma-
1976
11.
Alexanderson B: Pharmacokinetics of nontniptyline in man after single and multiple oral doses: the predictability of steady-state plasma concentrations from single-dose plasma-level data. Eur J Clin Pharmacol 4:82-91, 1972 12. Cooper TB, Allen D, Simpson GM: A sensitive method for the determination of amitniptyline and nortniptyline in human plasma. Psychopharmacological Communications 2:105-116,
1976
335
J Psychiatry
BRIEF
135:3, March
COMMUNICATIONS
Prediction BY
THOMAS
1978
of Individual B. COOPER,
M.A.,
Dosage AND
GEORGE
of Nortriptyline M.
SIMPSON,
The authors describe a technique that enables a physician to determine individual patient dosage requirements for nortriptylinefrom a single 24-hour blood sample. Because the technique reveals immediately those patients at the extremes of dosage ranges, toxicity and the needfor time-consuming titration ofthe dosage regimen can be avoided.
IN 1971 WE REPORTED that individual dosage requirements could be predicted from a single dose of lithium carbonate by analyzing blood samples collected at 8 discrete time points during the 24 hours following ingestion of the drug (1). In 1973 we reported a major modification of this procedure that enabled us to predict individual dosage requirements from a single blood sample collected 24 hours after a 600-mg oral dose oflithium carbonate (2). The application and usefulness of that procedure were described in a 2-year follow-up study at this center (3). Seifert and associates (4) used our basic concept in determining a suitable regression equation for a slow-release formulation, and Chang and associates (5) recently confirmed our original observation using the 25-hour rather than the 24-hour time point. The debate over whether there is a therapeutic threshold (lower limit below which patients do not improve) or therapeutic window (lower threshold below which and upper limit above which patients do not improve) for the plasma concentration of the various antidepressant and neuroleptic drugs is far from resolution (6). Of all the data published to date the most consistent and convincing, in our opinion, are those
M.B.,
CH.B.
for nortriptyline. In endogenous depression only, a therapeutic window of 50-150 ng/ml has been claimed and confirmed (7-9). Asbemg (10) has recently shown that in patients randomly assigned to either a 25-mg t.i.d. or 50-mg t.i.d. regimen of nomtriptyline, the majority of their steadystate plasma levels will be within the therapeutic range on a dosage of 50 mg t.i.d. ; however, many patients who receive 25 mg t.i.d. will be below the lower plasma threshold of 50 ng/ml. Asberg recommended that patients for whom blood level determinations are not available should be given 50 mg t.i.d. and those who do not respond should have their medication decreased rather than increased. Alexandemson (1 1) has shown the ability to predict steady-state’ plasma levels of nomtriptyline by determining the plasma clearance rate from a single oral dose ofnomtmiptyline. The correlation between this mate and the steady-state plasma level was r=.93. These data clearly indicate that this system is linear. Additional evidence for linearity is the very high correlation (r= .99) between the plasma clearance mate calculated for the single oral dose study and the same rates calculated from the steady-state data. This linearity in the system suggests that the technique we have used for lithium prediction would be applicable to the prediction of the steady-state plasma levels of nomtriptyline. In addition, the constancy of this prediction over time was clearly demonstrated when the prediction obtained from a single-dose curve was applied 2 years later to the same patients who were then given nortriptyline and had steady-state values determined (11).
EXPERIMENTAL Mr. Cooper is Chief Investigator, Clinical Psychopharmacology Laboratory, Rockland Research Institute, Orangebung, N.Y. 10962, where Dr. Simpson was Principal Research Psychiatrist. Dr. Simpson is now Professor of Psychiatry and Director, U.S.C.-Metropsychopharmacology Services, Norwalk, Calif. This work was supported tal Health Administration tute of Mental Health.
in part by Alcohol, grant MH-08240
The unteers
0002-953X/78/0003-0333$0.40
©
subjects were 18 physically healthy (10 women and 8 men) aged 21-51
Steady state is defined been receiving a fixed
Drug Abuse, and Menfrom the National Insti-
DESIGN
American
Psychiatric
volwho
operationally as the time when a patient has dosage of medication such that repeat blood
level determinations carried do not differ significantly.
1978
normal years,
out
Association
each
day
before
the
morning
dose
333
BRIEF
Am
COMMUNICATIONS
fully understood the project and gave informed consent. Subjects fasted for 12 hours and then ingested 50 mg of nortmiptyline by mouth. Blood samples were collected immediately before ingestion of the medication (t0) and at 1,2,3,4,6,8,12,24, and 36 hours after this dose. The volunteers were not allowed to lie down for the first hour after ingestion and were not allowed to eat until a light meal was served 4 hours after ingestion of the medication. After the initial single-dose expemiments, all subjects ingested nortmiptyline on a 25-mg b.i.d. (9:00 a.m. and 9:00 p.m.) regimen for a period of 15 days, and blood samples were collected on days 7,14,15, and 16 (immediately before the morning dose of medication). The samples were analyzed in duplicate for nortriptyline content using a gas-liquid chromatography (GLC) procedure and a nitrogen detector (12). This method has a coefficient of variation less than 5% at a level of5 ng/ml and can detect less than I ng/ml when a 3-mI sample is extracted.
AND
DISCUSSION
/35:3,
March
1978
TABLE 1 Correlation Between Individual Time Points from Single-Dose and Steady-State Plasma Levels* Hours
After
50-mg
Dose
Correlation
-.002 +194
3
+316
4 6
+.282
1
8
+535 +555
12 24 36
+.839 +918 +963
*All
subjects
received
25 mg
b.i.d.
Study
Coefficient
2
(r)
of nortriptyline.
FIGURE 1 Plot of 24-Hour Data Versus Steady-State Plasma Level of 18 Volunteers After Ingestion of 25 mg of Nortriptyline b.i.d.* 140-.
y= l6.6+4.3(x)
.E110-J LU > LU -J
RESULTS
J Psychiatry
-
90-
Ui
z
80-
-J
The steady-state plasma level of nortriptyline determined on days 13,14, and 15 were within 10% of the average value for each individual patient. With one exception all subjects had attained the steady-state condition by the seventh day of continuous medication. Alexanderson ( I 1 ) made a similar observation 6 days after treatment began. Correlations between the mean steady-state plasma level and individual time points of the initial singledose curve are shown in table 1 The strong correlations at 12, 24, and 36 hours make it possible to use any one of these time points for accurate prediction of the steady-state level. For convenience we have chosen to use the 24-hour time point (figure 1); i.e., the patient returns for the blood collection at the same time on the next day. These data, reduced to a simple table of prediction of dosage level, are shown in table 2. Inspection of table 2 shows drug dosage changes detemmined by a 1 ng/ml blood level difference. A method with a coefficient of variation of 5% at best can be expected to measure a true level of 20 ng/ml as anything from 18-22 ng/ml. The change of dosage at, for example, 19 ng/ml is calculated to take into account this variance. Thus a true value of 19 ng/ml determined as 21 ng/ml will result in levels within the therapeutic range whether a 50-mg b.i.d. or 25-mg t.i.d. regimen is followed. Analysis of these samples in duplicate will reduce this variance further and is recommended as a routine procedure. If one accepts that a therapeutic window exists for nortriptyline, then by using this technique one can predict within 48 hours (assuming 24-hour turnaround in the laboratory) the precise dosage requirements for each patient. Thus a patient is given 50-mg of nortriptyline by the physician and instructed to return to the physician’s office (or a laboratory) to have a blood
700
60x
0
-
z 40F-
xx
2010(-I)
.
334
5
10
24-HOUR
PLASMA
0 *r_
15
20
25
NORTRIPTYLINE
LEVEL
.92.
TABLE
2
Dosage Regimen Based on 24-Hour gestion of 50 mg of Nortriptyline Suggested
Blood
Level
(mg/mI)
Blood Levels After In-
Suggested
Dosage
Regimen
41* *The
30 (ng/ml)
mg
10 mg ranges
calculated
sample laboratory instructed (during a blood can then increased ed when be kept
are for
calculated
these
by extrapolation
experimental
data
ofthe (figure
regression
b.i.d.
bid. line
(y=a+bx)
1).
drawn at the same time the next day. If the is not open at that time, the patient can be to take the 50-mg dose at some other time the working hours of the laboratory) and have sample collected 24 hours later. The patient be started on medication immediately (e.g., to 150 mg/day) and the dosage can be adjustlaboratory data are available. The patient can on this adjusted regimen for a suitable period
Am
J Psychiatry
135:3,
March
BRIEF
1978
(4 weeks), and if therapeutic response is not achieved, theme would appear to be little point in changing an already optimum dosage; rather, a change in medication or treatment modality must be considered by the clinician.
COMMENT
REFERENCES I.
Bergner P-EE, Berniker K, Cooper TB, Ct al: Lithium kinetics in man: effect of variation in dosage pattern. Br J Pharmacol 49:328-339, 1973 2. Cooper TB, Bergnen P-EE, Simpson GM: The 24-hour serum lithium level as a prognosticator of dosage requirements. Am J Psychiatry 130:601-603, 1973 3. Cooper TB, Simpson GM: The 24-hour lithium level as a prognosticator of dosage requirements: a 2-year follow-up study.
Am J Psychiatry
We emphasize that this observation applies only to a serious endogenous depression. A relationship between drug plasma level and clinical efficacy in nonendogenous depression has not been established (6). Despite the paucity of good clear data on the efficacy of monitoring plasma levels of drugs, the recent spate of publications of analytical procedures for tncyclic drugs from clinical laboratories seems to indicate that major efforts are under way to establish laboratories for routine screening using these measures. Given this situation, it would seem that the technique we have described may be of considerable help in a practical clinical setting, i.e., bringing a patient into therapeutic mange rapidly.2 The physician does not have to titrate the patient with medication, and the occasional patients who metabolize a drug extremely slowly or rapidly, and who therefore require very different drug regimens from the average patient, can be identified immediately. These combined elements may result in precise tailoring of the medication to each individual patient’s needs, thus minimizing overmedication that could result in untoward side effects. 2
Preliminary study ranging
of
data this from
from
technique 95 to 128
four have nglml
patients
participating
in a prospective
achieved steady state on the dosage predicted.
blood
levels
COMMUNICATIONS
133:440-443,
1976
4.
Seifert R, Bremkamp H, Junge C: Vereinfachte lithiumeinstellung dench belastungstest. Psychopharmacologia 43:285-286, 1975 5. Chang 55, Pandey GN, Casper R, et al: Predicting the optimal lithium dosage. Proceedings of the American Society for Clinical Pharmacology and Therapeutics. Dallas, Tex, March 23-25, 1977 6. Cooper TB, Simpson GM, Lee JH: Thymoleptic and Neuroleptic Drug Plasma Levels in Psychiatry: Current Status in International Review ofNeurobiology. vol 19. New York, Academic Press, 1976, pp 269-309 7. Asbeng M. Cronholm B, Sjoqvist F. et al: Relationship between plasma level and therapeutic effect of nortniptyline. Br Med J
3:331-334, 1971 8. Knagh-Sonensen tniptyline levels 9.
10.
P. Asbeng M, Eggert-Hansen in endogenous depression.
1973 Ziegler yE. Clayton levels and therapeutic 463, 1976 Asberg M: Treatment macokinetic and
kopsychiatry
9:18-26,
PJ.
Taylor response.
of depression pharmacodynamic
C: Plasma norLancet 1:113-115,
JR. et al: Nortriptyline Clin Pharmacol Then with
tnicyclic aspects.
plasma 20:458-
drugs-phanPhanma-
1976
11.
Alexanderson B: Pharmacokinetics of nontniptyline in man after single and multiple oral doses: the predictability of steady-state plasma concentrations from single-dose plasma-level data. Eur J Clin Pharmacol 4:82-91, 1972 12. Cooper TB, Allen D, Simpson GM: A sensitive method for the determination of amitniptyline and nortniptyline in human plasma. Psychopharmacological Communications 2:105-116,
1976
335
