- Email: [email protected]
Penetration of corticosteroids into the lung: Evidence for a difference between methylprednisolone and prednisolone
Original
articles
Penetration of corticosteroids into the lung: Evidence for a difference between methylprednisolone and prednisolone Pakit Vichyanond, MD,* Charles G. Irvin, PhD, Gary L. Larsen, MD, Stanley J. Szefler, MD, and Malcolm R. Hill, PharmD Denver, CO/O Little is known about the penetration oj’corticosteroids. .su(.has methylprednisolone and prednisolone, into the lung, despite their common use in the treatment qf injlammatov lung disease., To compare methylprednisolone and prednisolone concentrations in the hronchoalveolar space, we administered these two c,orti~,o.steroici.sin a randomized, crossover fashion to 17 adult rabbits. A loading dose was administered and was ,followed by a continuous infifsion for 180 minutes to achieve steady-state plasma concentrutions between 200 to 2000 nglml. Serial playma samples were obtained during the ir$usions. Bronchoalveolar lavages (BAL) were performed at 180 minutes with sterile saline. Plasma and BAL jluid (BALF) were asx~yed for methylprednisolone and prednisolone concentration by high-performance liquid chromatography. Corticosteroid concentrations were normalized to urea concentrations in plurma and BALT. Generally, BALF corticosteroid concentration increased as plasma corlcentration increased. At plasma concentrations >800 ngl ml, BALF methylprednisolone corrcentrations increased e.uponentially, whereas the increase ,for prednisolone remained linear. BALF methylprednisolone was Jve times as high as that qj’prednisolone when plasma cot ticosteroid concentration was in the 2000 ng I ml range. With this continuous infusion technique, methy/prednisolone has u higher degree of‘ bronc~hoaiveolar penetration thun prednisolone, and these di$erences are greater at higher plasma concentrations. (J ALLERGY CL.'NIMMUNOL 1989;84:867-73.j
Corticosteroids are extre,mely important agents in the treatment of inflammatory lung disease. Despite the extensive use of corticosteroids in the treatment of disorders, such as asthma and interstitial pneumonitis, the ‘
I
I Abbreviation used
BALF: Bronchoalveolar lavage fluid i
I
the lung has been surprisingly limited. The alveolarcapillary unit consisting of vascular endothelial membranes, an interstitial space, and alveolar epithelium is analogous to the blood-brain barrier as a physiologic obstacle to drug delivery to a potential site of action. Braude and Rebuck’ were the first to examine this pharmacokinetic aspect for corticosteroids and reported that cortisol concentrations in the BALF to be similar to plasma concentrations. Later observation by the same investigators indicated that methylprednisolone was superior to prednisone in the ability to penetrate into the bronchoalveolar space.* These latter observations have received significant attention from investigators in the related fields, of whom some have expressed concerns in the interpretation of these results, since there was a marked heterogeneity in the disease conditions among the patients in the two treatment groups. Moreover. there was a disparity in the 867
888 Vichyanond et al.
J. ALLERGY
Drugs
TABLE I. Methylprednisolone and prednisolone dosing schedule and the targeted plasma corticosteroid concentration
Loading
dose*
(mglkd
0.13 0.33 0.75 1.00 1.33
Infusion rate* (mglkglhr)
CLIN. IMMUNOL. DECEMBER 1989
Target plasma concentration
hglml)
Prednisolone sodium succinate (Sigma Chemical Co., St. Louis, MO.) and commercially available methylprednisolone sodium succinate (Solu-Medrol; Upjohn Co., Kalamazoo, Mich.) were used. The doses were calculated as free alcohol corticosteroid and were dissolved in sterile saline for intravenous administration.
Experimental
protocol
0.13 0.33 0.75
200 500 1000
Corticosteroids were administered through the marginal ear vein via 23-gauge butterfly needle with a calibrated
1.00
1500
1.33
2000
22, Harvard Apparatus, South Natick, Mass.). Steady-state plasma corticosteroid concentrations in experimental animals were produced with a loading dose (bolus dose) and
*Loading doses and infusion rates were calculated from known rabbit pharmacokineticparametersto produce the desired target plasma concentration. Doses of the individual corticosteroids were calculated as the free alcohol.
dosing schemes and the routes of administration of the two corticosteroids. In addition, information on prednisolone concentrations (the active form of prednisone) were not available from this study. These findings have thus been regarded as incomplete observations and require further investigation. Pharmacokinetics of corticosteroids in the rabbit have been demonstrated to be remarkably similar to pharmacokinetics in man with regard to tissue distribution, metabolic clearance and pathways, metabolic interconversion, and protein-binding characteristics.3-5 Hence, the rabbit appears to be an ideal animal model for the study of various aspects of corticosteroid distribution. The purpose of this investigation was to study the delivery of systemically administered corticosteroids into the bronchoalveolar space and, in particular, we sought to determine whether there would be any significant difference in the diffusion of methylprednisolone or prednisolone into the BALF at similar and therapeutically achievable plasma concentrations. MATERIAL AND METHODS Animals and study design Adult New Zealand white rabbits, weighing between 3 and 4 kg, were used throughout the study. The Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and the National Institutes of Health was followed. These rabbits were studied in a randomized, crossover design, first to receive either methylprednisolone or prednisolone, and then crossed over to receive the alternate corticosteroid after a minimum 2- to 3-week interval. Since some animals did not survive the initial lavage experiments, 17 rabbits were required to complete the 28 experiments described below.
syringe infusion pump (Harvard syringe infusion pump No.
maintenancedose(by continuous infusion). The purposeof these doses was to produce the steady-state plasma corti-
costeroidconcentrationsin the range of 200 to 2000 ngiml (Table I) and were calculated by use of pharmacokinetic parametersin rabbits as previously described.6Serial blood sampleswere obtained through an indwelling catheter insertedinto the central ear artery of the contralateral ear at time 0, %, 1,2, and 3 hours. Plasmasampleswere separated immediately and stored at - 20” C until analysis. The animals were anesthetized with intramuscular xylazine (9 mg/kg) and ketamine (50 mg/kg) and were subjected to BAL 3 hours later, since it has been previously demonstrated that tissue uptake of corticosteroids were maximal at the end of the 3-hour continuous infusion.6 After endotracheal intubation with a 4-O endotracheal tube, a 6F double-lumen radiopaque Swan Ganz catheter (American Edwards Laboratories, Irvine, Calif.) was inserted into the right lower lobe. The position of the catheter tip was verified by a chest x-ray to be approximately 2 cm below the carina. BAL was performed with four successive aliquots of sterile saline (5 ml each). The total dwell time for lavage was minimized (to between 3 to 4 minutes) to reduce the potential for passive diffusion of drugs and urea from the vascular compartment into the lavage fluid.’ The percent recovery of these lavages were approximately 60% to 75% of the total.
Total leukocyte counts in BALF were determined immediately with a hemocytometer. Minimal quantities of red blood cells (two experiments for prednisolone and two for methylprednisolone) were lyzed with 2% acetic acid. The BALF cellular fraction was then pelleted at 130 g for IO minutes. This fraction was separated and reconstituted with 1 ml of normal saline for differential white cell count that was determined with regular Wright stain. No attempt was made to differentiate alveolar macrophages from large lymphocytes or monocytes. This group of cells was expressed as “mononuclear cells.” The remaining supernatant was frozen at -20” C until the time of analysis for corticosteroid and urea nitrogen concentration.
Corticosteroid
analysis
Plasma and BALF corticosteroid concentrations were analyzed with a previously described high-performance liquid
VOLUME NUMBER
Corticosteroid
84 6. PART
penetration
into the lung
869
Prednisolone
Prednisone
1
2
3
77MEAFTER LOADINGDOSE (HOURS) FIG. 1. Resultant plasma prednisolone and prednisone (inactive metabolite) concentrations are illustrated for a representative experiment after infusion of prednisolone. The dosing scheme outlined in Table I was applied to achieve and maintain steady-state plasma prednisolone concentrations for 180 minutes before lavage. In this experiment, a loading dose of 0.75 mgikg was followed by a maintenance dose of 0.75 mgikgihr.
chromatography technique* with silica gel column (Zorbax Sil; DuPont Chemistry, Wilmington, Del.) as a solid phase and a mobile phase containing 400 : 200: 350 : 1.5methylene chloride/ethylene dichloride/ heptane/glacial acetic acid.’ Ethanol was added in varying quantities (22 to 40 ml/L of mobile phase) to adjust the retention time for methylprednisolone and prednisolone between 8 to 12 minutes. The intraday and interday coefficient of variation of this system is <50/n, and the lower limit for steroid detection was 5 rig/ml.
Urea nitrogen
concentration
analysis
Since BALF represents diluted epithelial lining fluid of varying degrelz, normalization of BALF corticosteroid concentration is required for further comparative purposes. This is accomplished by dividing BALF cotticosteroid concentration with concentration of a macromolecule that readily diffuses across alveolar capillary plasma membrane. Urea has been demonstrated to be an excellent agent for this purpose and was used in this study.’ Similarly, plasma corticosteroid concentrations were corrected with plasma urea concentrations. Measurement of BALF and plasma urea was performed by a spectrophotometric assay (blood urea nitrogen, end point 20, Sigma Chemical Co.). To compensate for a different range of urea concentrtions, sample/reagent ratios for pla:jma and BALF were 1:25 and 1 :2, respectively. Optical density was read at 340 nm by a Beckman DU 6 spectrophotometer (Beckman Instruments, Irvine, Calif.).
Data analysis All data are expressed as mean + SEM. Cell counts and urea (plasma and BALF) were compared between treatment groups with the unpaired Student’s t test. The relationship between BALF and plasma cotticosteroid concentrations
(corrected with urea nitrogen concentrations) were examined with linear and various polynomial regression models.
RESULTS With the dosing schedule as presented in Table I, a plateau for plasma corticosteroid concentrations was achieved within 30 minutes, and these concentrations were maintained throughout the 3-hour experimental period (Fig. 1). For prednisolone (active corticosteroid) infusions, both prednisolone and prednisone (inactive metabolite of prednisolone) could be adequately measured in the rabbit plasma and BALF samples. The ratios of prednisoloneiprednisone in the plasma of these rabbits varied from threefold to tenfold (mean t SEM of 6.4 + 0.61). The mean prednisolonelprednisone ratio in BALF was 2.6 -t 1.2, indicating a higher relative concentration of prednisone in BALF as compared to plasma. The ratio of plasma methylprednisolone/ methylprednisone was higher and more inconsistent (range, 10 to 63). The total number of leukocytes recovered from BALF was similar in both groups (Fig. 2). Moreover, mononuclear cells were the predominant cells in both corticosteroid groups. No polymorphonuclear leukocytes were observed. This suggested that there was an absence of inflammation at the time of lavage and that the condition
of the lung at the site of lavage
was similar between the two corticosteroid-treatment groups. Additionally, the number of leukocytes recovered were similar to the number recovered from rabbits not receiving corticosteroids. ” For urea analysis, it was determined that the upper limit of detection of the blood urea nitrogen increases
870
Vichyanond
et al.
J. ALLERGY
Total
Leukocyte
MPn
Count
Pn
Mononuclear
FIG. 2. (MPnIof cells normal
Cell
MPn
la)
CLIN. IMMUNOL. DECEMBER 1989
Pn ib)
Cellularity of BALF. There were no significant differences between methylprednisolone and prednisolone (P&treated groups in either the total leukocyte count (a) or the type observed in the BALF. Most cells were mononuclear cells, which is characteristic for rabbits. Data presented represent mean ? SEM of 14experiments on each corticosteroid.
as the sample to reagent ratio decreases (less sample volume). Conversely, the accuracy of detection at lower concentration range (< 10 kg/ml) improves with increasing sample/reagent ratio (larger sample volume). Therefore, different ratios were selected for plasma and BALF urea concentration (1: 25 for plasma and 1: 2 for BALF). The percent coefficient of variation of this system at these ratios was 5.6% for plasma urea and 2.6% for BALF urea. Mean t SEM urea for methylprednisolone versus prednisolone, respectively, was 4.05 t 0.53 p,g/ml versus 3.08 t 0.38 pg/ml (BALF) and 200.5 2 8.46 pg/ml versus 201.6 2 8.57 kg/ml (plasma). No statistical differences were noted between the two corticosteroid-treatment groups for plasma or BALF urea (p > 0.05). The relationship between the plasma and BALF corticosteroid concentrations is illustrated in Fig. 3. It is apparent that when the plasma corticosteroid concentrations were 15 ng/fl.g (or actual plasma corticosteroid concentration of 5 1000 rig/ml), no significant difference in BALF concentrations between the two corticosteroids were observed. However, at plasma concentrations 25 ng/pg, BALF methylprednisolone began to increase progressively, whereas BALF prednisolone remained relatively linear. For example, at plasma corticosteroid concentration of 10 rig/kg (range of actual concentration between 1500 to 2000 ng / ml), BALF methylprednisolone concentration was five times as high as BALF prednisolone. Although there is a trend for this difference to continue to be greater, we did not extend the study beyond the plasma corticosteroid concentrations > 1500 to 2000 rig/ml, since this range of concentrations is not therapeutically achievable in most clinical circumstances. It is interesting to note that at the plasma concentration range of 1500 to 2000 rig/ml, corrected BALF meth-
ylprednisolone concentration was approximately five times as high as plasma methylprednisolone concentration, whereas BALF prednisolone concentration was equal to plasma concentration. The best curve fit for methylprednisolone was a second degree polynomial (r = 0.95; p < O.OOl), whereas that of prednisolone was of linear fit (r = 0.54; p < 0.05). DISCUSSION The results of our study convincingly demonstrated a clear difference in the amount of corticosteroid recovered from the bronchoalveolar space between methylprednisolone and prednisolone when these were administered as continuous infusion. This difference was more pronounced at higher plasma concentrations and existed despite achieving similar concentrations of both corticosteroids systemically. As a result, to achieve similar corticosteroid concentrations in the bronchoalveolar space, systemic doses of prednisolone would have to be at least several times greater. Understanding the barriers of antibiotic transport to the central nervous system has greatly improved the therapeutic approach in the treatment of patients with meningitis.“’ I2 Similarly, it follows that there could be a therapeutic advantage to treat pulmonary disorders with drugs that demonstrate a higher penetrability into the lung. In a previous study of Braude and Rebuck,* plasma prednisone concentrations were observed to be in the lower end of the pharmacologic range, and plasma methylprednisolone concentrations extended to include the entire range observed during acute intravenous dosing (e.g., status asthmaticus). Moreover, information on prednisolone, the active metabolite of prednisone, was not reported. In our study, plasma corticosteroid concentrations of prednisolone and
VOLUME NUMBER
Corticosteroid
84 6. PART 1
-. 0
5
10
penetration
into the lung
871
15
Plasma Concentration (ng corticosteroidhg urea) FIG. 3. Correlation
between corrected BALF and corrected plasma steroid significant difference between methylprednisolone and prednisolone BALF observed at higher plasma steroid concentrations with methylprednisolone nisolone at this range. The best curve fit for methylprednisolone was a second (r = 0.95; p < 0.001) and that for prednisolone was a linear fit (I’ = 0.54; p
methylprednisolone were comparable and encompassed the en.tire range of clinically achievable concentrations within the pharmacologic range of interest. We compared concentrations of the active corticosteroids, methylprednisolone and prednisolone ( Fig. 3). Our findings not only substantiated the observations of Braude and Rebuck’ but also indicated that a significant difference of the two corticosteroids in BALF occurred at higher plasma concentrations. Plasma corticosteroid concentrations of 1000 to 2000 p.g/ ml (approximately 5 to 10 ng of steroid per microgram of urea) may be clinically relevant in situations when patients are .receiving short-course, large-dose corticosteroids (60 to 80 mg orally) or when they are receiving large-dose intermittent therapy for status asthmaticus or other inflammatory lung diseases, such as idiopathic pulmonary fibrosis.‘” Because saturation of prednisolone tissue uptake in the rabbit occurs after 3 hours of continuous intravenous infusion,6 BAL was performed at this time point. Tissue uptake of prednisolone in the homogenized rabbit ILungwas significant compared to unbound plasma prednisolone concentration, although this uptake was not as high as uptake observed in the intestine and in the liver of the same animals.6 It is possible that the pulmonary tissue uptake, as previously analyzed with lung homogenates, reflects the combination of vascular and tissue compartments of the lung. Interconversion of prednisolone (active) to prednisone (inactive) occurred in rabbit lung, but not to the degree as interconversion observed in the liver.6 The ratios of active to inactive corticosteroids observed in this investigation were somewhat different than in pre-
concentrations. A concentrations was greater than preddegree polynomial < 0.05).
vious studies. The prednisolone I prednisone ratio in plasma in our study was 6.4 Z!I 0.6. This ratio decreased to 2.6 f 1.2 in BALF. A previous investigation with a similar model and infusion scheme’ demonstrated a similar ratio in homogenized lung and in plasma. A difference in the prednisolone / prednisone ratio in BALF and homogenized lung could be due to residual blood or plasma in the homogenized lung tissue. This could alternatively be explained by an active interconversion process in the bronchoalveolar space between prednisolone and prednisone. favoring the formation of the inactive metabolite, prednisone. Unfortunately, accurate ratios for methylprednisolone and the inactive species, methylprednisone, could not be derived. This is perhaps because methylprednisone coelutes with corticosterone (native rabbit glucocorticoid). This interferes with the measurement of methylprednisone, despite suppression of corticosterone with prolonged administration of exogenous corticosteroid.4 In addition, BALF has several unidentified lipid components that interfere with methylprednisone and, rarely, prednisone peaks. Therefore, BALF and plasma methylprednisone concentrations are not reliably measured with our current highperformance liquid chromatography system in the rabbit. Similar data in human lung tissue are not available, although interconversion is likely present, since cortisone is converted to cortisol in fetal lung tissue. I4 The possibility that differences in solubility of methylprednisolone and prednisolone in saline used for BAL or application of urea as a dilutional marker account for the observed differences in BALF corticoc;teroid concentrations are considerations. However,
872 Vichyanond et al.
both corticosteroids are directly soluble in saline in the concentration range of 25 to 250 rig/ml, which was encounteredover the range of dosesused in this study (data not presented).In thesesolubility studies, virtually 100%of the intended concentrationsof both prednisolone and methylprednisolone were recovered from saline. Therefore, relative insolubility of prednisolone in saline does not account for the higher BALF methylprednisolone concentrations that were observed.Urea is a useful markerfor dilutional factors associatedwith the BAL technique.’ It is susceptible to the effect of dwell time, since urea may diffuse from the plasma into the instilled saline.15We have used urea to facilitate comparison between prednisolone and methylprednisolone and not for estimation of the amount of epithelial lining fluid recovered.Additionally, the effect of dwell time was minimized since all lavages were completed within 3 to 4 minutes. Therefore, use of urea or the particular lavagetiming sequenceapplied in this investigation doesnot artifactually account for the differences in BALF corticosteroid concentrationsthat we observed. The bronchoalveolar spacewith macrophagesas its major cellular constituent is a relevant site of action for corticosteroids in the treatment of inflammatory lung diseases. Stimulated macrophages produce a wide array of mediatorsandchemoattractantsfor other inflammatory cells, such as neutrophils and eosinophils . 16-19 Moreover, the number of glucocorticoid receptorson alveolar macrophageswere found to be approximately twofold to threefold higher than blood leukocytes, indicating high avidity of macrophages for glucocorticoids.2~The main anti-inflammatory action of glucocorticoids is believed to be the antagonistic effect on phospholipaseA, via the production of lipomodulin or lipocortin.” Thus, glucocorticoids may not only reduce the production of mediators of inflammation but also reduce the transfer of cellular componentsof inflammation, such as eosinophils and neutrophils to the inflammatory site in the lung. Since the anti-inflammatory effectsof corticosteroidsshould increase as their concentrations in BALF increase, penetration of theseagentsinto the lung is important. Several differences in pharmacokinetic properties between methylprednisolone and prednisolone have been observed that could explain the present observations.22Greater volume of distribution of methylprednisolone along with its longer mean residence time and higher octanol/water gradient (higher lipid solubility) favor its tissue delivery over prednisolone. This is especially true in tissuesthat require diffusion of drugs acrossmembranebarriers that are presentin the alveolar-capillary system. The protein binding characteristicsof the two corticosteroids are also dif-
J. ALLERGY
CLIN. IMMUNOL. DECEMBER 1989
ferent.3-5Methylprednisolone largely binds with low affinity and high capacity to albumin. Prednisolone binds preferentially with high affinity and low capacity to corticosteroid-binding globulin or transcortin. This provides constant(linear) free concentrationsof methylprednisolone over the entire achievable range of plasmaconcentrations,whereasconcentrationsof free prednisoloneincreasesin a nonlinear fashion once the binding capacity of transcortin has been exceeded (higher free plasma prednisolone at plasma concentration of prednisolone >200 rig/ml). With this differencein protein-binding characteristics,BALF prednisolone would be anticipatedto be greaterthan BALF methylprednisolone at higher plasma concentrations becauseof the higher free fraction of prednisolone than methylprednisolone. Our findings suggest that this may not be the only mechanismfor the retention of corticosteroids in BALF. Several other possible explanations for this observation include differences in the rate of diffusion of binding proteins for the two corticosteroidsinto bronchoalveolar space,differences in the lung uptake for the different corticosteroids, different rates of metabolic interconversion by lung cells, and the possibility of the presenceof an active secretion mechanismfor methylprednisolone, but not prednisolone. Diffusion of transcortin and albumin into the bronchoalveolar space could occur at different rates and indirectly result in the difference between the two corticosteroid concentrations. These pharmacokinetic aspects have not been critically examined elsewhere to our knowledge. Minor alterations in the chemical structureof a substance could result in a major difference in tissue uptake in the lung. An example of this phenomenon is the complete lung uptake of norepinephrine after intravenous infusion in contrast to the absenceof absorption for epinephrine administered via the same route.23Another example was the significant absorption of inhaled atropine sulfate into systemic circulation observed while minimal amount of atropine methylnitrate was measuredin the blood stream.24It may be possible that the addition of a methyl group to the C-6 position of the corticosteroid molecule for methylprednisolone results in significant retention of this steroid in the lung and possibly in other tissues. It has been demonstrated,for example, that methylprednisolone administered into bovine knee remains in the joint fluid for at least 3 months.25It is also possible that the rate of metabolic interconversion of methylprednisolonein the lung may be different than prednisolone. Finally, the presenceof an active secretion mechanismfor various types of corticosteroids from the vascular compartment to different tissue
VOLUME NUMBER
Corticosteroid
84 6. PART 1
compartments could explarn the differences observed in our investigation. We are presently conducting several other pharmacokinetic investigations, including single-dose studies to delineate the mechanisms to explain the difference in lung penetration between the two corticosteroids. ‘The results of such investigations would be clinically applicable, since glucocorticoids are generally administered in single doses. With the currently available and future knowledge of corticosteroid pharmacokinetics. the use of corticosteroids will be optimized in different disease conditions with less potential for serious adverse effects. We thank Helga W. Cole for manuscript preparation and Eleanor Brown, MT, ASCP, for technical assistance. REFERENCES I. Braude AC. Rebuck AS. Pulmonary disposition of cortisol. Ann Intern Med 1982;97:59. 2. Braude AC, Rehuck AS. Prednisone and methylprednisolone disposition in the lung. Lancet 1983;2:995. 3. Rocci ML. Jusko WJ. Dose-dependent protein binding and disposition of prednisolone in rabbits. J Pharm Sci 1981: 79:1201. 4. Ebllng WF’. Szefler SJ. Jbsko WJ. Methylprednisolone disposition in rabbits. Analysis, prodrug conversion, reversible metabolism. and comparison with man. Drug Metab Dispos 1985;13:296. 5. Eblinp WY, Milsap RI, Szefler SJ, Jusko WJ. 6-alphamethylprednisolone and 6-alpha-methylprednisone plasma protein binding in human and rabbit. J Pharm Sci 1986;75:760. 6. Khalafallah N, Jusko WJ. Tissue distribution of prednisolone in rabbit. 1 Pharmacol Exp Ther 1984;229:719. 7. Rennard SI, Basset G, Lecossier D. O’Donnell K, Pinkston P. Martin PG. Estimation of volume of epithelial lining fluid recovered 3y lavage using lurea as a marker for dilution. J Appl Physiol 1986;60:532. 8. Ebling WI:, SzeAer SJ. Jusko WJ. Analysis of cortisol. methylprednisolone, and nethylprednisolone hemisucccinate Absence of effects of troleandomycin on ester hydrolysis. J Chromatogr 1984;305:271-80. 9. Bartoszek M. Brenner AM, Szefler SJ. Prednisolone and methylprednisolone kinetics ir. children receiving anticonvulsant therapy. C’lin Pharmacol Ther 1987;42:424. 10. Murphy KR. Wilson MC, Irvin CG, et al. The requirement for polyrrorphonuclear leukocytes in the late asthmatic re-
II
12 13
I‘l
15
I6 17 18
19
20
21
22
23
24
25
penetration
into the lung? 873
sponse and heightened airways reactivity m ,111animal model Am Rev Respir Dis 1986:134:62-8. Klein JO. Feigin RD, McCracken GH. Report of the task force on diagnosis and management of mcninpiti\ PcdiatricI 1986: 78(suppl):970. Stutman HR. Marks MI Therapy for bacterial mcnmgitis. Which drugs and for hou long’? J Pcdiatr IOX’i:l lU:Xi?. Keogh BA. Bernard OJ. Hunninghake GW. l.mc BR, Price DL, Crystal RG. Effect of internuttent high-dose parcnteral corticoatermds on the alvcolitis of idiopathic pulmonnrv fibrohis. Am Rev Respir Dis 1983:127:18-12. Murphy BEP. Ontogeny uf cortisol-cortisone mtcrconversion m human tissuc5: a role in human feta! devcltrpment. J Steroid Biochcm 1981:14:81 I. Marcy TW. Merrill WW. Rumkin JA, Re~nulda HY Llmlrations of using urea to quantify epithelial lining fiuid recovered by bronchoalveolar lavagc. Am Rev Respir Dis I987;135: 1276 Morley J. Role of prostaglandins secreted by macrophages in the infiammatory process. Lymphokines I98 I :-1:377 Fels A. Cohn ZA. The alveolar macrophagc J Appl Physiol 1986:60:353. Amoux B. Simeos-Caeiro MH, Landes A, klathicu M, Duroux P, Benveniste J. Alveolar macrophages from asthmatic patients release platelet-activating factor (Paf-acether) and lyso Pafacether when stimulated with the specilic allergen Am Rev Respir Dis 1982;125:A50. Joseph M, Tonne1 AB, Torpier G, Capron ,I. .4moux B. Benveniate J. lnvolvement of immunoglobulin E in the secretory process o! alveolar macrophages from asthmatic oatients. J Clin Invest 1983;7 1:211. Ozaki T. Yasuoka S. Nakayama T. Tsubura E. Glucocorticoid receptors in human alveolar macrophages and peripheral blood cells. Clin Exp Immunol 1982;47:505. Hirata F, Schiffman E, Venkatasubramanian K, Salomen D, Axelod J. A phospholipase AL inhibitory protein in rabbit neutrophils induced by glucocorticoids. Proc Nat1 ,Zcad Sci USA 1980;77:2533. SzeRer SJ. Ebling WF, Georgltis JW. Jusko WJ. Methylprednisolone versus prednisolone pharmacokinetlcs in relation to dose in adults. Eur J Clin Phannacol 1986:30:323. Alabaster VA. Inactivatron of endogenous ammes in the lungs. In: Bakhle YS. Vane JR. eds. Metabolic functlom of the lungs. New York: Marcel Dekker, 1977:3. Gross NJ. Skorodin MS. Anticholinergic agents. In: Jenne JW, Murphy S. eds. Drug therapy for asthma. Research andclinical Practice. New York: Marcel Dekker. 19X7:615. Toutain PL, Alvinerie M. Fayolle P, Ruckebush Y. Bovine plasma and synovial kinetics of methylpredmsolone and methylprednisolone acetate after intra-articular administration of methylprednisolone acetate. J Pharm Exp Ther 19X6:236:794.
articles
Penetration of corticosteroids into the lung: Evidence for a difference between methylprednisolone and prednisolone Pakit Vichyanond, MD,* Charles G. Irvin, PhD, Gary L. Larsen, MD, Stanley J. Szefler, MD, and Malcolm R. Hill, PharmD Denver, CO/O Little is known about the penetration oj’corticosteroids. .su(.has methylprednisolone and prednisolone, into the lung, despite their common use in the treatment qf injlammatov lung disease., To compare methylprednisolone and prednisolone concentrations in the hronchoalveolar space, we administered these two c,orti~,o.steroici.sin a randomized, crossover fashion to 17 adult rabbits. A loading dose was administered and was ,followed by a continuous infifsion for 180 minutes to achieve steady-state plasma concentrutions between 200 to 2000 nglml. Serial playma samples were obtained during the ir$usions. Bronchoalveolar lavages (BAL) were performed at 180 minutes with sterile saline. Plasma and BAL jluid (BALF) were asx~yed for methylprednisolone and prednisolone concentration by high-performance liquid chromatography. Corticosteroid concentrations were normalized to urea concentrations in plurma and BALT. Generally, BALF corticosteroid concentration increased as plasma corlcentration increased. At plasma concentrations >800 ngl ml, BALF methylprednisolone corrcentrations increased e.uponentially, whereas the increase ,for prednisolone remained linear. BALF methylprednisolone was Jve times as high as that qj’prednisolone when plasma cot ticosteroid concentration was in the 2000 ng I ml range. With this continuous infusion technique, methy/prednisolone has u higher degree of‘ bronc~hoaiveolar penetration thun prednisolone, and these di$erences are greater at higher plasma concentrations. (J ALLERGY CL.'NIMMUNOL 1989;84:867-73.j
Corticosteroids are extre,mely important agents in the treatment of inflammatory lung disease. Despite the extensive use of corticosteroids in the treatment of disorders, such as asthma and interstitial pneumonitis, the ‘
I
I Abbreviation used
BALF: Bronchoalveolar lavage fluid i
I
the lung has been surprisingly limited. The alveolarcapillary unit consisting of vascular endothelial membranes, an interstitial space, and alveolar epithelium is analogous to the blood-brain barrier as a physiologic obstacle to drug delivery to a potential site of action. Braude and Rebuck’ were the first to examine this pharmacokinetic aspect for corticosteroids and reported that cortisol concentrations in the BALF to be similar to plasma concentrations. Later observation by the same investigators indicated that methylprednisolone was superior to prednisone in the ability to penetrate into the bronchoalveolar space.* These latter observations have received significant attention from investigators in the related fields, of whom some have expressed concerns in the interpretation of these results, since there was a marked heterogeneity in the disease conditions among the patients in the two treatment groups. Moreover. there was a disparity in the 867
888 Vichyanond et al.
J. ALLERGY
Drugs
TABLE I. Methylprednisolone and prednisolone dosing schedule and the targeted plasma corticosteroid concentration
Loading
dose*
(mglkd
0.13 0.33 0.75 1.00 1.33
Infusion rate* (mglkglhr)
CLIN. IMMUNOL. DECEMBER 1989
Target plasma concentration
hglml)
Prednisolone sodium succinate (Sigma Chemical Co., St. Louis, MO.) and commercially available methylprednisolone sodium succinate (Solu-Medrol; Upjohn Co., Kalamazoo, Mich.) were used. The doses were calculated as free alcohol corticosteroid and were dissolved in sterile saline for intravenous administration.
Experimental
protocol
0.13 0.33 0.75
200 500 1000
Corticosteroids were administered through the marginal ear vein via 23-gauge butterfly needle with a calibrated
1.00
1500
1.33
2000
22, Harvard Apparatus, South Natick, Mass.). Steady-state plasma corticosteroid concentrations in experimental animals were produced with a loading dose (bolus dose) and
*Loading doses and infusion rates were calculated from known rabbit pharmacokineticparametersto produce the desired target plasma concentration. Doses of the individual corticosteroids were calculated as the free alcohol.
dosing schemes and the routes of administration of the two corticosteroids. In addition, information on prednisolone concentrations (the active form of prednisone) were not available from this study. These findings have thus been regarded as incomplete observations and require further investigation. Pharmacokinetics of corticosteroids in the rabbit have been demonstrated to be remarkably similar to pharmacokinetics in man with regard to tissue distribution, metabolic clearance and pathways, metabolic interconversion, and protein-binding characteristics.3-5 Hence, the rabbit appears to be an ideal animal model for the study of various aspects of corticosteroid distribution. The purpose of this investigation was to study the delivery of systemically administered corticosteroids into the bronchoalveolar space and, in particular, we sought to determine whether there would be any significant difference in the diffusion of methylprednisolone or prednisolone into the BALF at similar and therapeutically achievable plasma concentrations. MATERIAL AND METHODS Animals and study design Adult New Zealand white rabbits, weighing between 3 and 4 kg, were used throughout the study. The Guide for the Care and Use of Laboratory Animals prepared by the National Academy of Sciences and the National Institutes of Health was followed. These rabbits were studied in a randomized, crossover design, first to receive either methylprednisolone or prednisolone, and then crossed over to receive the alternate corticosteroid after a minimum 2- to 3-week interval. Since some animals did not survive the initial lavage experiments, 17 rabbits were required to complete the 28 experiments described below.
syringe infusion pump (Harvard syringe infusion pump No.
maintenancedose(by continuous infusion). The purposeof these doses was to produce the steady-state plasma corti-
costeroidconcentrationsin the range of 200 to 2000 ngiml (Table I) and were calculated by use of pharmacokinetic parametersin rabbits as previously described.6Serial blood sampleswere obtained through an indwelling catheter insertedinto the central ear artery of the contralateral ear at time 0, %, 1,2, and 3 hours. Plasmasampleswere separated immediately and stored at - 20” C until analysis. The animals were anesthetized with intramuscular xylazine (9 mg/kg) and ketamine (50 mg/kg) and were subjected to BAL 3 hours later, since it has been previously demonstrated that tissue uptake of corticosteroids were maximal at the end of the 3-hour continuous infusion.6 After endotracheal intubation with a 4-O endotracheal tube, a 6F double-lumen radiopaque Swan Ganz catheter (American Edwards Laboratories, Irvine, Calif.) was inserted into the right lower lobe. The position of the catheter tip was verified by a chest x-ray to be approximately 2 cm below the carina. BAL was performed with four successive aliquots of sterile saline (5 ml each). The total dwell time for lavage was minimized (to between 3 to 4 minutes) to reduce the potential for passive diffusion of drugs and urea from the vascular compartment into the lavage fluid.’ The percent recovery of these lavages were approximately 60% to 75% of the total.
Total leukocyte counts in BALF were determined immediately with a hemocytometer. Minimal quantities of red blood cells (two experiments for prednisolone and two for methylprednisolone) were lyzed with 2% acetic acid. The BALF cellular fraction was then pelleted at 130 g for IO minutes. This fraction was separated and reconstituted with 1 ml of normal saline for differential white cell count that was determined with regular Wright stain. No attempt was made to differentiate alveolar macrophages from large lymphocytes or monocytes. This group of cells was expressed as “mononuclear cells.” The remaining supernatant was frozen at -20” C until the time of analysis for corticosteroid and urea nitrogen concentration.
Corticosteroid
analysis
Plasma and BALF corticosteroid concentrations were analyzed with a previously described high-performance liquid
VOLUME NUMBER
Corticosteroid
84 6. PART
penetration
into the lung
869
Prednisolone
Prednisone
1
2
3
77MEAFTER LOADINGDOSE (HOURS) FIG. 1. Resultant plasma prednisolone and prednisone (inactive metabolite) concentrations are illustrated for a representative experiment after infusion of prednisolone. The dosing scheme outlined in Table I was applied to achieve and maintain steady-state plasma prednisolone concentrations for 180 minutes before lavage. In this experiment, a loading dose of 0.75 mgikg was followed by a maintenance dose of 0.75 mgikgihr.
chromatography technique* with silica gel column (Zorbax Sil; DuPont Chemistry, Wilmington, Del.) as a solid phase and a mobile phase containing 400 : 200: 350 : 1.5methylene chloride/ethylene dichloride/ heptane/glacial acetic acid.’ Ethanol was added in varying quantities (22 to 40 ml/L of mobile phase) to adjust the retention time for methylprednisolone and prednisolone between 8 to 12 minutes. The intraday and interday coefficient of variation of this system is <50/n, and the lower limit for steroid detection was 5 rig/ml.
Urea nitrogen
concentration
analysis
Since BALF represents diluted epithelial lining fluid of varying degrelz, normalization of BALF corticosteroid concentration is required for further comparative purposes. This is accomplished by dividing BALF cotticosteroid concentration with concentration of a macromolecule that readily diffuses across alveolar capillary plasma membrane. Urea has been demonstrated to be an excellent agent for this purpose and was used in this study.’ Similarly, plasma corticosteroid concentrations were corrected with plasma urea concentrations. Measurement of BALF and plasma urea was performed by a spectrophotometric assay (blood urea nitrogen, end point 20, Sigma Chemical Co.). To compensate for a different range of urea concentrtions, sample/reagent ratios for pla:jma and BALF were 1:25 and 1 :2, respectively. Optical density was read at 340 nm by a Beckman DU 6 spectrophotometer (Beckman Instruments, Irvine, Calif.).
Data analysis All data are expressed as mean + SEM. Cell counts and urea (plasma and BALF) were compared between treatment groups with the unpaired Student’s t test. The relationship between BALF and plasma cotticosteroid concentrations
(corrected with urea nitrogen concentrations) were examined with linear and various polynomial regression models.
RESULTS With the dosing schedule as presented in Table I, a plateau for plasma corticosteroid concentrations was achieved within 30 minutes, and these concentrations were maintained throughout the 3-hour experimental period (Fig. 1). For prednisolone (active corticosteroid) infusions, both prednisolone and prednisone (inactive metabolite of prednisolone) could be adequately measured in the rabbit plasma and BALF samples. The ratios of prednisoloneiprednisone in the plasma of these rabbits varied from threefold to tenfold (mean t SEM of 6.4 + 0.61). The mean prednisolonelprednisone ratio in BALF was 2.6 -t 1.2, indicating a higher relative concentration of prednisone in BALF as compared to plasma. The ratio of plasma methylprednisolone/ methylprednisone was higher and more inconsistent (range, 10 to 63). The total number of leukocytes recovered from BALF was similar in both groups (Fig. 2). Moreover, mononuclear cells were the predominant cells in both corticosteroid groups. No polymorphonuclear leukocytes were observed. This suggested that there was an absence of inflammation at the time of lavage and that the condition
of the lung at the site of lavage
was similar between the two corticosteroid-treatment groups. Additionally, the number of leukocytes recovered were similar to the number recovered from rabbits not receiving corticosteroids. ” For urea analysis, it was determined that the upper limit of detection of the blood urea nitrogen increases
870
Vichyanond
et al.
J. ALLERGY
Total
Leukocyte
MPn
Count
Pn
Mononuclear
FIG. 2. (MPnIof cells normal
Cell
MPn
la)
CLIN. IMMUNOL. DECEMBER 1989
Pn ib)
Cellularity of BALF. There were no significant differences between methylprednisolone and prednisolone (P&treated groups in either the total leukocyte count (a) or the type observed in the BALF. Most cells were mononuclear cells, which is characteristic for rabbits. Data presented represent mean ? SEM of 14experiments on each corticosteroid.
as the sample to reagent ratio decreases (less sample volume). Conversely, the accuracy of detection at lower concentration range (< 10 kg/ml) improves with increasing sample/reagent ratio (larger sample volume). Therefore, different ratios were selected for plasma and BALF urea concentration (1: 25 for plasma and 1: 2 for BALF). The percent coefficient of variation of this system at these ratios was 5.6% for plasma urea and 2.6% for BALF urea. Mean t SEM urea for methylprednisolone versus prednisolone, respectively, was 4.05 t 0.53 p,g/ml versus 3.08 t 0.38 pg/ml (BALF) and 200.5 2 8.46 pg/ml versus 201.6 2 8.57 kg/ml (plasma). No statistical differences were noted between the two corticosteroid-treatment groups for plasma or BALF urea (p > 0.05). The relationship between the plasma and BALF corticosteroid concentrations is illustrated in Fig. 3. It is apparent that when the plasma corticosteroid concentrations were 15 ng/fl.g (or actual plasma corticosteroid concentration of 5 1000 rig/ml), no significant difference in BALF concentrations between the two corticosteroids were observed. However, at plasma concentrations 25 ng/pg, BALF methylprednisolone began to increase progressively, whereas BALF prednisolone remained relatively linear. For example, at plasma corticosteroid concentration of 10 rig/kg (range of actual concentration between 1500 to 2000 ng / ml), BALF methylprednisolone concentration was five times as high as BALF prednisolone. Although there is a trend for this difference to continue to be greater, we did not extend the study beyond the plasma corticosteroid concentrations > 1500 to 2000 rig/ml, since this range of concentrations is not therapeutically achievable in most clinical circumstances. It is interesting to note that at the plasma concentration range of 1500 to 2000 rig/ml, corrected BALF meth-
ylprednisolone concentration was approximately five times as high as plasma methylprednisolone concentration, whereas BALF prednisolone concentration was equal to plasma concentration. The best curve fit for methylprednisolone was a second degree polynomial (r = 0.95; p < O.OOl), whereas that of prednisolone was of linear fit (r = 0.54; p < 0.05). DISCUSSION The results of our study convincingly demonstrated a clear difference in the amount of corticosteroid recovered from the bronchoalveolar space between methylprednisolone and prednisolone when these were administered as continuous infusion. This difference was more pronounced at higher plasma concentrations and existed despite achieving similar concentrations of both corticosteroids systemically. As a result, to achieve similar corticosteroid concentrations in the bronchoalveolar space, systemic doses of prednisolone would have to be at least several times greater. Understanding the barriers of antibiotic transport to the central nervous system has greatly improved the therapeutic approach in the treatment of patients with meningitis.“’ I2 Similarly, it follows that there could be a therapeutic advantage to treat pulmonary disorders with drugs that demonstrate a higher penetrability into the lung. In a previous study of Braude and Rebuck,* plasma prednisone concentrations were observed to be in the lower end of the pharmacologic range, and plasma methylprednisolone concentrations extended to include the entire range observed during acute intravenous dosing (e.g., status asthmaticus). Moreover, information on prednisolone, the active metabolite of prednisone, was not reported. In our study, plasma corticosteroid concentrations of prednisolone and
VOLUME NUMBER
Corticosteroid
84 6. PART 1
-. 0
5
10
penetration
into the lung
871
15
Plasma Concentration (ng corticosteroidhg urea) FIG. 3. Correlation
between corrected BALF and corrected plasma steroid significant difference between methylprednisolone and prednisolone BALF observed at higher plasma steroid concentrations with methylprednisolone nisolone at this range. The best curve fit for methylprednisolone was a second (r = 0.95; p < 0.001) and that for prednisolone was a linear fit (I’ = 0.54; p
methylprednisolone were comparable and encompassed the en.tire range of clinically achievable concentrations within the pharmacologic range of interest. We compared concentrations of the active corticosteroids, methylprednisolone and prednisolone ( Fig. 3). Our findings not only substantiated the observations of Braude and Rebuck’ but also indicated that a significant difference of the two corticosteroids in BALF occurred at higher plasma concentrations. Plasma corticosteroid concentrations of 1000 to 2000 p.g/ ml (approximately 5 to 10 ng of steroid per microgram of urea) may be clinically relevant in situations when patients are .receiving short-course, large-dose corticosteroids (60 to 80 mg orally) or when they are receiving large-dose intermittent therapy for status asthmaticus or other inflammatory lung diseases, such as idiopathic pulmonary fibrosis.‘” Because saturation of prednisolone tissue uptake in the rabbit occurs after 3 hours of continuous intravenous infusion,6 BAL was performed at this time point. Tissue uptake of prednisolone in the homogenized rabbit ILungwas significant compared to unbound plasma prednisolone concentration, although this uptake was not as high as uptake observed in the intestine and in the liver of the same animals.6 It is possible that the pulmonary tissue uptake, as previously analyzed with lung homogenates, reflects the combination of vascular and tissue compartments of the lung. Interconversion of prednisolone (active) to prednisone (inactive) occurred in rabbit lung, but not to the degree as interconversion observed in the liver.6 The ratios of active to inactive corticosteroids observed in this investigation were somewhat different than in pre-
concentrations. A concentrations was greater than preddegree polynomial < 0.05).
vious studies. The prednisolone I prednisone ratio in plasma in our study was 6.4 Z!I 0.6. This ratio decreased to 2.6 f 1.2 in BALF. A previous investigation with a similar model and infusion scheme’ demonstrated a similar ratio in homogenized lung and in plasma. A difference in the prednisolone / prednisone ratio in BALF and homogenized lung could be due to residual blood or plasma in the homogenized lung tissue. This could alternatively be explained by an active interconversion process in the bronchoalveolar space between prednisolone and prednisone. favoring the formation of the inactive metabolite, prednisone. Unfortunately, accurate ratios for methylprednisolone and the inactive species, methylprednisone, could not be derived. This is perhaps because methylprednisone coelutes with corticosterone (native rabbit glucocorticoid). This interferes with the measurement of methylprednisone, despite suppression of corticosterone with prolonged administration of exogenous corticosteroid.4 In addition, BALF has several unidentified lipid components that interfere with methylprednisone and, rarely, prednisone peaks. Therefore, BALF and plasma methylprednisone concentrations are not reliably measured with our current highperformance liquid chromatography system in the rabbit. Similar data in human lung tissue are not available, although interconversion is likely present, since cortisone is converted to cortisol in fetal lung tissue. I4 The possibility that differences in solubility of methylprednisolone and prednisolone in saline used for BAL or application of urea as a dilutional marker account for the observed differences in BALF corticoc;teroid concentrations are considerations. However,
872 Vichyanond et al.
both corticosteroids are directly soluble in saline in the concentration range of 25 to 250 rig/ml, which was encounteredover the range of dosesused in this study (data not presented).In thesesolubility studies, virtually 100%of the intended concentrationsof both prednisolone and methylprednisolone were recovered from saline. Therefore, relative insolubility of prednisolone in saline does not account for the higher BALF methylprednisolone concentrations that were observed.Urea is a useful markerfor dilutional factors associatedwith the BAL technique.’ It is susceptible to the effect of dwell time, since urea may diffuse from the plasma into the instilled saline.15We have used urea to facilitate comparison between prednisolone and methylprednisolone and not for estimation of the amount of epithelial lining fluid recovered.Additionally, the effect of dwell time was minimized since all lavages were completed within 3 to 4 minutes. Therefore, use of urea or the particular lavagetiming sequenceapplied in this investigation doesnot artifactually account for the differences in BALF corticosteroid concentrationsthat we observed. The bronchoalveolar spacewith macrophagesas its major cellular constituent is a relevant site of action for corticosteroids in the treatment of inflammatory lung diseases. Stimulated macrophages produce a wide array of mediatorsandchemoattractantsfor other inflammatory cells, such as neutrophils and eosinophils . 16-19 Moreover, the number of glucocorticoid receptorson alveolar macrophageswere found to be approximately twofold to threefold higher than blood leukocytes, indicating high avidity of macrophages for glucocorticoids.2~The main anti-inflammatory action of glucocorticoids is believed to be the antagonistic effect on phospholipaseA, via the production of lipomodulin or lipocortin.” Thus, glucocorticoids may not only reduce the production of mediators of inflammation but also reduce the transfer of cellular componentsof inflammation, such as eosinophils and neutrophils to the inflammatory site in the lung. Since the anti-inflammatory effectsof corticosteroidsshould increase as their concentrations in BALF increase, penetration of theseagentsinto the lung is important. Several differences in pharmacokinetic properties between methylprednisolone and prednisolone have been observed that could explain the present observations.22Greater volume of distribution of methylprednisolone along with its longer mean residence time and higher octanol/water gradient (higher lipid solubility) favor its tissue delivery over prednisolone. This is especially true in tissuesthat require diffusion of drugs acrossmembranebarriers that are presentin the alveolar-capillary system. The protein binding characteristicsof the two corticosteroids are also dif-
J. ALLERGY
CLIN. IMMUNOL. DECEMBER 1989
ferent.3-5Methylprednisolone largely binds with low affinity and high capacity to albumin. Prednisolone binds preferentially with high affinity and low capacity to corticosteroid-binding globulin or transcortin. This provides constant(linear) free concentrationsof methylprednisolone over the entire achievable range of plasmaconcentrations,whereasconcentrationsof free prednisoloneincreasesin a nonlinear fashion once the binding capacity of transcortin has been exceeded (higher free plasma prednisolone at plasma concentration of prednisolone >200 rig/ml). With this differencein protein-binding characteristics,BALF prednisolone would be anticipatedto be greaterthan BALF methylprednisolone at higher plasma concentrations becauseof the higher free fraction of prednisolone than methylprednisolone. Our findings suggest that this may not be the only mechanismfor the retention of corticosteroids in BALF. Several other possible explanations for this observation include differences in the rate of diffusion of binding proteins for the two corticosteroidsinto bronchoalveolar space,differences in the lung uptake for the different corticosteroids, different rates of metabolic interconversion by lung cells, and the possibility of the presenceof an active secretion mechanismfor methylprednisolone, but not prednisolone. Diffusion of transcortin and albumin into the bronchoalveolar space could occur at different rates and indirectly result in the difference between the two corticosteroid concentrations. These pharmacokinetic aspects have not been critically examined elsewhere to our knowledge. Minor alterations in the chemical structureof a substance could result in a major difference in tissue uptake in the lung. An example of this phenomenon is the complete lung uptake of norepinephrine after intravenous infusion in contrast to the absenceof absorption for epinephrine administered via the same route.23Another example was the significant absorption of inhaled atropine sulfate into systemic circulation observed while minimal amount of atropine methylnitrate was measuredin the blood stream.24It may be possible that the addition of a methyl group to the C-6 position of the corticosteroid molecule for methylprednisolone results in significant retention of this steroid in the lung and possibly in other tissues. It has been demonstrated,for example, that methylprednisolone administered into bovine knee remains in the joint fluid for at least 3 months.25It is also possible that the rate of metabolic interconversion of methylprednisolonein the lung may be different than prednisolone. Finally, the presenceof an active secretion mechanismfor various types of corticosteroids from the vascular compartment to different tissue
VOLUME NUMBER
Corticosteroid
84 6. PART 1
compartments could explarn the differences observed in our investigation. We are presently conducting several other pharmacokinetic investigations, including single-dose studies to delineate the mechanisms to explain the difference in lung penetration between the two corticosteroids. ‘The results of such investigations would be clinically applicable, since glucocorticoids are generally administered in single doses. With the currently available and future knowledge of corticosteroid pharmacokinetics. the use of corticosteroids will be optimized in different disease conditions with less potential for serious adverse effects. We thank Helga W. Cole for manuscript preparation and Eleanor Brown, MT, ASCP, for technical assistance. REFERENCES I. Braude AC. Rebuck AS. Pulmonary disposition of cortisol. Ann Intern Med 1982;97:59. 2. Braude AC, Rehuck AS. Prednisone and methylprednisolone disposition in the lung. Lancet 1983;2:995. 3. Rocci ML. Jusko WJ. Dose-dependent protein binding and disposition of prednisolone in rabbits. J Pharm Sci 1981: 79:1201. 4. Ebllng WF’. Szefler SJ. Jbsko WJ. Methylprednisolone disposition in rabbits. Analysis, prodrug conversion, reversible metabolism. and comparison with man. Drug Metab Dispos 1985;13:296. 5. Eblinp WY, Milsap RI, Szefler SJ, Jusko WJ. 6-alphamethylprednisolone and 6-alpha-methylprednisone plasma protein binding in human and rabbit. J Pharm Sci 1986;75:760. 6. Khalafallah N, Jusko WJ. Tissue distribution of prednisolone in rabbit. 1 Pharmacol Exp Ther 1984;229:719. 7. Rennard SI, Basset G, Lecossier D. O’Donnell K, Pinkston P. Martin PG. Estimation of volume of epithelial lining fluid recovered 3y lavage using lurea as a marker for dilution. J Appl Physiol 1986;60:532. 8. Ebling WI:, SzeAer SJ. Jusko WJ. Analysis of cortisol. methylprednisolone, and nethylprednisolone hemisucccinate Absence of effects of troleandomycin on ester hydrolysis. J Chromatogr 1984;305:271-80. 9. Bartoszek M. Brenner AM, Szefler SJ. Prednisolone and methylprednisolone kinetics ir. children receiving anticonvulsant therapy. C’lin Pharmacol Ther 1987;42:424. 10. Murphy KR. Wilson MC, Irvin CG, et al. The requirement for polyrrorphonuclear leukocytes in the late asthmatic re-
II
12 13
I‘l
15
I6 17 18
19
20
21
22
23
24
25
penetration
into the lung? 873
sponse and heightened airways reactivity m ,111animal model Am Rev Respir Dis 1986:134:62-8. Klein JO. Feigin RD, McCracken GH. Report of the task force on diagnosis and management of mcninpiti\ PcdiatricI 1986: 78(suppl):970. Stutman HR. Marks MI Therapy for bacterial mcnmgitis. Which drugs and for hou long’? J Pcdiatr IOX’i:l lU:Xi?. Keogh BA. Bernard OJ. Hunninghake GW. l.mc BR, Price DL, Crystal RG. Effect of internuttent high-dose parcnteral corticoatermds on the alvcolitis of idiopathic pulmonnrv fibrohis. Am Rev Respir Dis 1983:127:18-12. Murphy BEP. Ontogeny uf cortisol-cortisone mtcrconversion m human tissuc5: a role in human feta! devcltrpment. J Steroid Biochcm 1981:14:81 I. Marcy TW. Merrill WW. Rumkin JA, Re~nulda HY Llmlrations of using urea to quantify epithelial lining fiuid recovered by bronchoalveolar lavagc. Am Rev Respir Dis I987;135: 1276 Morley J. Role of prostaglandins secreted by macrophages in the infiammatory process. Lymphokines I98 I :-1:377 Fels A. Cohn ZA. The alveolar macrophagc J Appl Physiol 1986:60:353. Amoux B. Simeos-Caeiro MH, Landes A, klathicu M, Duroux P, Benveniste J. Alveolar macrophages from asthmatic patients release platelet-activating factor (Paf-acether) and lyso Pafacether when stimulated with the specilic allergen Am Rev Respir Dis 1982;125:A50. Joseph M, Tonne1 AB, Torpier G, Capron ,I. .4moux B. Benveniate J. lnvolvement of immunoglobulin E in the secretory process o! alveolar macrophages from asthmatic oatients. J Clin Invest 1983;7 1:211. Ozaki T. Yasuoka S. Nakayama T. Tsubura E. Glucocorticoid receptors in human alveolar macrophages and peripheral blood cells. Clin Exp Immunol 1982;47:505. Hirata F, Schiffman E, Venkatasubramanian K, Salomen D, Axelod J. A phospholipase AL inhibitory protein in rabbit neutrophils induced by glucocorticoids. Proc Nat1 ,Zcad Sci USA 1980;77:2533. SzeRer SJ. Ebling WF, Georgltis JW. Jusko WJ. Methylprednisolone versus prednisolone pharmacokinetlcs in relation to dose in adults. Eur J Clin Phannacol 1986:30:323. Alabaster VA. Inactivatron of endogenous ammes in the lungs. In: Bakhle YS. Vane JR. eds. Metabolic functlom of the lungs. New York: Marcel Dekker, 1977:3. Gross NJ. Skorodin MS. Anticholinergic agents. In: Jenne JW, Murphy S. eds. Drug therapy for asthma. Research andclinical Practice. New York: Marcel Dekker. 19X7:615. Toutain PL, Alvinerie M. Fayolle P, Ruckebush Y. Bovine plasma and synovial kinetics of methylpredmsolone and methylprednisolone acetate after intra-articular administration of methylprednisolone acetate. J Pharm Exp Ther 19X6:236:794.