BRIEF REPORT
`
`91
`
`PHARMACOKINETICS OF METHOTREXATE ADMINISTERED BY
`INTRAMUSCULAR AND SUBCUTANEOUS INJECTIONS IN PATIENTS
`WITH RHEUMATOID ARTHRITIS
`
`PAUL J. BROOKS, WILLIAM J. SPRUILL, ROY C. PARISH, and DANIEL A. BIRCHMORE
`
`The serum concentrations and the pharmacoki·
`netics of low-dose methotrexate (MTX) were compared
`after both intramuscular
`(IM) and subcutaneous (SQ)
`injections in 5 patients with rheumato id arthritis. Val(cid:173)
`ues for the observed peak concentration, the time to the
`observed peak concentration, and the area under the
`time versus concentration curve for IM injections were
`not significantly different from these values for SQ
`injections. These results suggest that IM and SQ are
`interchangeable routes of administration. SQ adminis(cid:173)
`tration may be a more convenient and less painful way
`of administering low-dose MTX .
`
`(MTX), a folic acid antagonist,
`Methotrexate
`has recently been approved by the Food and Drug
`Administration for use in patients with severe rheuma(cid:173)
`toid arthritis that is refractory to conventional therapy.
`The proposed beneficial effect of MTX in treating
`rheumatic diseases is its ability to inhibit inflammatory
`synovial cell turnover. decrease exudation in the joint
`spaces, and impair the response to histamine and other
`(1-4). Treatment bas centered
`vasoactive substances
`
`From the Department of Pharmacy Practice , The Univer·
`sity of Georgia College of Pharmacy , Athens.
`Supported by a grant from The Upjohn Company.
`Paul J. Brooks, PharmD: Fellow in Adult Internal Medicine
`Pharmacy Practice (currently Clinical Assistant Professor of Phar(cid:173)
`macy Practice, The Univers ity of North Carolina School of Phar(cid:173)
`macy at Fayetteville Area Health Education Center); William J.
`Spruill, PharmD : Associate Professor of Pharmacy Practice; Roy C.
`Parish, PharmD: Assistant Professor of Pharmacy Practice; Daniel
`A Birchmorc. MD: Athens , GA .
`Address reprint requests to Paul J. Brooks. PharmD. Fay(cid:173)
`etteville Area Health Education Center, 1601-8 Owen Drive. Fay(cid:173)
`etteville. NC 28304.
`Submitted for publication May 18, 1989; accepted in revised
`form August 14, J989.
`
`Arthritis and Rheumatism, Vol. 33, No. J (January 1990)
`
`in
`around the use of very low doses administered
`weekly intervals by oral (PO), intravenous (IV), and
`intramuscular (IM) routes (3,5-9).
`The intramuscular route is a desirab le choice
`for parenteral drug administration because of the com(cid:173)
`pleteness of absorption relative to the oral route, peak
`concentrations
`that are similar to those achieved using
`the IV route, and slower drug absorption and pro(cid:173)
`longed exposure to the drug compared with IV(cid:173)
`administered MTX (3,6-8). As an alternative method
`of administration, subcutaneous
`(SQ) injections may
`also exhibit these beneficial pharmacokinetic patterns
`and would have the potential advantages of patient
`self-administration at home and greater patient com(cid:173)
`fort than with weekly IM injection s given in the
`physician's office. fn the present study, we compared
`the serum concentrations and the pharmacokinetic
`parameters of MTX after IM and SQ administration
`in
`patients with rheumatoid arthritis.
`Patients and methods. The stu dy population
`consisted of 5 patients (age range 45-75 years) who
`had severe rheumatoid arthritis and were currently
`receiving MTX (Table 1). All patients were under the
`care of a board-certified rheumatologist and had expe(cid:173)
`rienced an unsatisfactory
`response
`to nonsteroidal
`antiinflammatory drugs and intramuscular gold ther(cid:173)
`apy. The patients had no history of hepatic disease,
`alcoholism, active peptic ulcer disease, or renal insuf(cid:173)
`ficiency . Patients who required additional antiinflam(cid:173)
`matory medication were permitted to continue taking
`their medication while receiving MTX .
`I week
`Each patient received 2 treatments,
`apart, given in a randomly assigned order. One treat(cid:173)
`ment consisted of the patient's usual dose adminis(cid:173)
`tered IM Oateral midthigh); the other treatment was
`
`Page 1 of 4
`
`KOIOS Exhibit 1008
`
`

`
`92
`
`BRIEF REPORTS
`
`Table 1. Characteristics of the 5 rheumatoid arthritis patients
`studied
`
`Patient
`
`Age/sex
`
`Weight (kg)
`
`I
`2
`3
`4
`5
`
`45/F
`65/F
`75/F
`53/M
`56/M
`
`61
`80
`90
`75
`66
`
`Methotrexate
`dose (mg)*
`
`25.0
`15.0
`12.5
`25.0
`20.0
`
`• This dose was given on 2 occasions 1 week apart: intramuscularly
`(lM) followed by subcutaneously (SQ) I week later, or SQ followed
`by IM I week later.
`
`the same dose administered SQ (lateral upper arm). An
`indwelling venous cannula was used to collect serial
`blood samples from each patient at O (baseline), 0.25,
`0.50, 0.75, 1.0, 1.5, 2.0, 4.0, and 8.0 hours after the
`injection. Blood was allowed to clot , and the serum
`was separated and stored at -10°C . The serum MTX
`concentration was determined in duplicate by a fluo(cid:173)
`rescence polarization immunoassay technique using
`the Abbott TDx clinical analyzer (lO) (Abbott Labo(cid:173)
`ratories, North Chicago, IL). This analyzer is reported
`to have a sensitivity of 0.01 moles/liter, and has
`coefficients of variation within assays and between
`assays of 8.09% and 9.20%, respeciively , for the
`0.07-moles/liter contro l concentration, and 3.94% and
`5.15%, respectively, for the 5.0-moles/liter control
`concentration. Assay cross-reactivity of 7-hydroxy(cid:173)
`methotrexate is reported as 1.5% (10).
`Time versus conce ntration data for each patient
`receiving each treatment were fitted to the appropriate
`exponential
`pharmacokinetic model , using
`the
`RSTRIP pharmacokinetic computer software (Micro(cid:173)
`Math, Salt Lake City, UT). Response variables exam(cid:173)
`ined included
`the observed peak concentration
`(Cmax), time to the observed peak concentration
`
`(Tmax), area under the time versus concentration
`curve (AUC), and the elimination (ke) and absorption
`(ka) rate constants. The AUC was calculated using the
`trapezoidal rule. The rate constants ke and ka were
`estimated by iterative least-squares methods using the
`RSTRTP software. Cmax and AUC values were nor(cid:173)
`malized for the dose, since patients received doses
`titrated to individual response, and are reported as
`Cmax/dose and AUC/dose, respectively.
`The statistical significance of the observed dif(cid:173)
`ferences in the pharmacokinetic data after administra(cid:173)
`tion by the different routes was evaluated using the
`paired-difference !-test for the response variables ke,
`ka, Cmax/dose, and AUC/dose. The Wilcoxon
`matched pairs signed rank test was used for differ(cid:173)
`ences in Tmax because it is unlikely that time is
`normally distributed. P values less than 0.05 were
`considered significant.
`Results. Pharmacokinetic data for the IM and
`SQ routes of MTX administration are shown in Table
`2. Values for the Cmax/dose were variable. Peak
`concentration data from the same patient after the 2
`routes of administration showed that the drug concen(cid:173)
`trations were higher after the IM dose in 2 patients,
`higher after the SQ dose in 1 patient, and equivalent in
`2 patients. The peak concentration (Tmax) occurred
`sooner and the rate of absorption (ka) was faster after
`the SQ injection in 4 of 5 patients. Percent differences
`in AUC/dose measurements after SQ and IM injec(cid:173)
`tions were 5% for patients 3 and 4, 14% for patients 2
`and 5, and 25% fot patient 1. The elimination rate
`constant (ke) was variable, and ranged from 0.14
`hours -• to 0.33 hours -• after the SQ doses, and 0.22
`hours - • to 0.34 hours- 1 after the IM doses.
`Statistical data regarding the null hypothe sis
`(that the mean difference between treatments for each
`
`Table 2. Pharmacokinetic data comparing intramuscular (IM) and subcutaneous (SQ) administration of methotrexate in 5 rheumatoid arthritis
`patients*
`
`Ke (hours - 1)
`
`Ka (hours - ')
`
`Cmax/dose MTX
`(µmo les/liter x
`mg)
`
`Tmax (hours)
`
`AUC/dose MTX
`(µmoles x
`hours/liter x mg)
`
`Patient
`
`IM
`
`SQ
`
`IM
`
`SQ
`26.75
`0.08
`0.23
`0.53
`5.67
`0.33
`0.23
`J
`0.48
`0.36
`0.08
`5.24
`0.10
`0.07
`2.72
`0.20
`0.34
`0.37
`0.43
`0.70
`1.17
`2
`0.29
`0.30
`5.44
`0.08
`0.09
`3.42
`3
`0.41
`0.39
`0.50
`1.03
`0.41
`5.36
`0.07
`1.50
`1.58
`0.25
`0.27
`0.39
`1.25
`0.o7
`4
`2.19
`0.12
`5
`0.61
`2.00
`0.25
`0.07
`35.30
`0.14
`0.22
`0.71
`"' Ke = elimination rate constant; Ka = absorption rate constant; Cmax = observed peak concentration; MTX == methotrexate; Tmax == time
`to the observed peak concentration ; AUC = area under the time versus concentration curve.
`
`IM
`
`SQ
`
`IM
`
`SQ
`
`IM
`
`SQ
`
`Page 2 of 4
`
`KOIOS Exhibit 1008
`
`

`
`BRIEF REPORTS
`
`93
`
`Table 3. Stati stical analysis of the differences in the pharmacoki (cid:173)
`netic data of the IM and SQ routes of MTX adm inistrat ion in 5
`rheumatoid arthritis patients•
`
`Response
`variable
`
`Difference t
`
`p
`
`20% difference 30% difference
`
`Power to dete ct
`
`Ke
`Ka
`Cmax
`AUC
`Tmax
`
`0.03 ± 0.09
`0.74 ± 19.80
`- 0.01 ± 0.03
`0.03 ± 0.08
`O.o2
`
`0.49
`0.94
`0.27
`0.37
`>0.05
`
`<0 .5:j:
`<0 .5:j:
`0.88§
`0.61:j:
`ND
`
`<0.5 :j:
`<0.5:j:
`0.99§
`0.90§
`ND
`
`• ND = not determined ; se,e Table 2 for other definitions .
`t IM - SQ. Tmax value is the median ; other values are the mean ±
`SD.
`:j: Difference between IM value and SQ value was not large enough
`to enable rejection of the null hypothesis.
`§ Null hypothesis accepted .
`
`is 0) are shown in Table 3. Calculated P
`parameter
`values exceeded
`the significance value of 0.05 for
`every response variable . To estimate the possibility of
`a Type II statistical error (i.e., falsely accepting the
`null hypothesis) , an analysis of power was performed
`to determine the power of the tests to detect clinically
`important differences at the 0.05 significance level.
`The power to detect a =::20% difference in Cmax was
`0.88, and the power to detect a ~30% difference in the
`AUC was 0 .90. Using these results from the power
`analysis , the null hypothesis for differences in Cmax
`and AUC was accepted. However, there was insuffi(cid:173)
`cient statistical evidence to either reject or accept the
`null hypothesis for the other parameters (Table 3).
`Discussion. Several studies have compared the
`pharmacokinetics of MTX by the IV , IM, and PO
`routes of administration
`(6,7,11). MTX administered
`by injection has been shown to produce higher serum
`concentrations
`and more complete absorption
`than
`does orally administered MTX. Specifically, intramus(cid:173)
`cularly administered MTX resulted in rapid and com(cid:173)
`plete absorption and in higher serum concentrations
`than did oral administration,
`and it provided peak
`concentrations similar to those observed following IV
`administration . Balis et al (12) compared pharmacoki(cid:173)
`netic data obtained after low doses of MTX were
`administered
`subcutaneously
`and orally
`to rhesus
`monkeys and to children with lymphoblastic leukemia.
`Those authors concluded that SQ administration was a
`feasible way to deliver MTX because
`it was well
`tolerated , efficiently absorbed , and it overcame prob(cid:173)
`lems of variable absorption seen after oral dosing (12).
`The resuJts of this study suggest that the SQ
`route achieves
`serum concentration
`versus
`time
`
`curves similar to the IM route. Statistical analysis
`suggests that the pharmacokinetic parameters are sim(cid:173)
`ilar for these 2 routes of administration . No statisti(cid:173)
`cally significant differences were observed for any
`response variable . However , an acceptable analysis of
`power value of 80% was reached for the variables
`Cmax and AUC, but not for the variables ke and ka.
`Thus, undetected differences in ke and ka may exist.
`Although changes in ke should not be dependent upon
`the administration
`technique, differences in ke would
`not be unexpected, since samples were taken I week
`apart, and intrasubject variability after drug therapy is
`not uncommon.
`The ka values showed considerable variability.
`The absorption rate was more rapid after SQ injection
`than after IM injection in all but I patient, whose rate
`of absorption was more rapid after IM administration.
`It is interesting to note that this patient had very little
`muscle mass , which may have affected the absorption
`rate. Slight differences in absorption rates (ka) would
`be expected when changing drug administration sites.
`Other possible factors altering
`the absorption
`rate
`include changes in the injection technique and differ(cid:173)
`ences in the distribution of blood circulation at dif(cid:173)
`ferent time s.
`has
`The metabolite 7-hydroxymethotrexate
`displayed significant blood concentration during me(cid:173)
`tabolism and rriay contribute
`to the clinical effect of
`methotrex.ate (5). However,
`the concentration of this
`metabolite was not determined in this study, because
`its formation should not influence drug absorption.
`The sampling interval of 8 hours seemed appro(cid:173)
`priate because 'it exceeded 2 drug half-lives in every
`case , and the drng concentrations during the 8-hour
`sample period 'P,proximated
`the limits of detection of
`the assay . Tfle ;least-squares approach used to calcu(cid:173)
`late ke and ka 4tilizes information from all data points
`to calculate the' optimal fit of the function to the data ;
`this eliminates ·th_e need for observations over several
`drug half-lives.
`Although patient acceptance was not assessed
`as part of this investigation, no patients complained of
`problems asso¢1a~ed with SQ administration , and most
`patients reporte,d :that the SQ injection was less painful
`than the IM iajection.
`These findings suggest that MTX concentra(cid:173)
`tions achieved by each method of delivery are statis(cid:173)
`tically and clinically similar, and that IM and SQ
`injections are interchangeable routes of MTX adminis(cid:173)
`tration . Although_ this study is considered preliminary
`because of the small sample size , our data support the
`
`Page 3 of 4
`
`KOIOS Exhibit 1008
`
`

`
`94
`
`BRIEF REPORTS
`
`routine use of subcutaneous MTX administration.
`allowing flexibility in the treatment of rheumatoid
`arthritis.
`
`REFERENCES
`
`I. Healey LA: The current status of methotrexate use in
`rheumatic diseases. Bull Rheum Dis 36: 1-10, 1986
`2. Hurd ER : Immunosuppressive
`and antiinflammatory
`propertie s of cyclophosphamide , azathioprine and meth(cid:173)
`otrexate . Arthritis Rheum 16:84-88 , L973
`3. Lentendre PW, DeJong DJ, Miller DR: The use of
`in rheumatoid arthritis . Drug Intcll Clin
`methotrexate
`Pharm 19:349-358 , 1985
`4. Steinberg AD: An approach
`to the use of immunosup(cid:173)
`pressive drugs in nonmalignant diseases. J Allergy Clio
`Immunol 52:242-250, 1973
`5. Evan s WE, Crom WR , Yalow ich JC : Methotrexate,
`Applied Pharmacokinetics . Second edition. Edited by
`WE Evans, JJ Schentag, WJ Jusko. Spokane , Applied
`Therapeutics, 1986
`6. Freeman-Narrod M, Gerstley BJ, Engstrom PF: Com(cid:173)
`parison of serum concentrations of methotrexate after
`
`various routes of admini stration . Cancer 36: 1619- 1624,
`1975
`7. Kamel RS , Al-Hakiem MH, Rademaker M, Meyrick
`Thomas RH, Munro DD: Pharmacokinetics of small
`doses of methotrexate
`in patients with psoriasis. Acta
`Derm Venereol (Stockh) 68:267- 270, 1988
`8. Michaels RM, Nashe! DJ , Leonard A , Sliwinski AJ,
`Derbes SJ: Weekly intravenous methotrexate
`in the
`treatment of rheumatoid arthr itis. Arthritis Rheum 25 :
`339-341 , 1982
`9. Teresi ME , Crom WR, Choi KE, Mirro J, Evans WE:
`Methotrexate bioavailability after oral and intramuscu(cid:173)
`lar administration
`in children . J Pediatr 110:788-792,
`1987
`10. Abbott Laboratories Diagnostics Division: TDX System
`Assays Manual. North Chicago , IL , Abbott Laborato(cid:173)
`ries , 1984
`11. Campbell MA, Perrier DG , Dorr RT, Alberts DS, Finley
`PR: Methotrexate : bioavailability and pharmacokinet(cid:173)
`ics. Cancer Treat Rep 69:833-838, 1985
`12. Balis FM, Mirro J , Reaman GH, Evans WE, McCully C,
`Doherty KM, Murphy RF, Jeffries S , Poplack DG:
`Pharmacokinetics of subcutaneous methotrexate . J Clin
`Oncol 6: 1882-1886, 1988
`
`Page 4 of 4
`
`KOIOS Exhibit 1008