Correlation between the drug penetration and the blanching effect of topically applied hydrocortisone creams in human beings

Correlation between the drug penetration and the blanching effect of topically applied hydrocortisone creams in human beings Daniele Caron, PhD,a Catherine Queille-Roussel, MD,b Vinod P. Shah, PhD,c and Hans Schaefer, PhDa Va/bonne and Nice, France, and Rockville, Maryland Two topical formulations of 2.5% hydrocortisone were tested to investigate the possible correlation between the blanching response of topically applied hydrocortisone and its release from the formulation after its penetration into the statum corneum. After application of the cream, the stratum corneum was removed at different intervals and the hydrocortisone content was measured by radioimmunoassay. The blanching responses (expressedas mean score vs time) showed a 1.5-fold difference (significant at p = 0.0026) between the two creams at the maximum effect, which was reached for both creams after 16 hours. The amounts of hydrocortisone present in the stratum corneum after every application period showed a twofold difference (p == 0.0001) between the two creams. However,in contrast to the blanching, these quantities remained constant from 4 to 24 hours.This steady state characteristic may be partialIy due to the use of occlusion and to the excess formulation remaining at the surface. Although there is no parallelism with time between the results of the two methods, both give clear evidence of a difference in stratum corneum penetration and pharmacologic activity of the two formulations. Drug penetrations and pharmacodynamic activities of the two creams can be correlated with in vitro release characteristics. (J AMACAD DERMATOl 1990;23:45862.)

A major problem in the development of a generic product is the determination of the bioequivalence with a formulation already on the market. Both pharmacodynamic and pharmacokinetic assessment can be made to evaluate bioequivalence. The blanching effect of topically applied corticosteroids is a pharmacologic activity that correlates well with clinical efficacy': 2 and that can be used as a measure of bioequivalence and local bioavailability. The aim of the present study was to compare blanching (pharmacodynamic activity) with bioavailability (penetration), as measured by drug level in the skin.! MATERIAL AND METHODS The commercial 2.5% hydrocortisone formulations used were Synacort cream (cream A) and Hytone cream (cream B). The two creams were coded to provide blind conditions. From the Centre International deRecherches Dermatologiques, Sophia Antipolis, Valbonne," theCentre de Pharmacologic Clinique Appliquee a la Dermatologie, Hopital Pasteur, Nice," and the Office of Research Resources, Center for Drug Evaluation and Research, Food and Drug Administration; Rockville." Accepted for publication Dec. 29, 19&9. Reprint requests: Daniele Caron, PhD, Centre International de Recherches Dermatologiques, Sophia Antipolis, 06560 Valbonne, France. 16/1/19368

458

Twelvehealthy white consentingvolunteers (nine men; three women) participated in the study. The test site was the upper part of the back.

Drug penetration study Methodology. For the drug penetration study, 25 loLl of formulation was applied toa 3 em?(2 X 1.5ern) area. The area was occluded with aluminum foil, covered with polyethylenefilm,and maintained in place with Blenderm tape (3M Medical-Surgical Div., St. Paul, Minn.). To determine the amount of drug penetrated in the stratum corneum (sq, as a function of time, six occlusion time intervals, 4, 6, 8, 16,20, and 24 hours, were selected. At the end of each application period the occlusivecover was removed and the area was wiped clean of excess formulation. Stripping. One hour after removal, each zone was stripped 10 times successively with a 3 cm 2 cellophane tape (Tesafilm, 2 em width; BDFTesa, Le Plessy Robinson). A special roller was used to provide constant pressure (670 gm/cm-) during stripping. The first stripping was consideredto contain excess(nonpenetrated) formulation. Determination of hydrocortisonelevels intheceUophane tape strippings. Each strip was stirred for 30 minutes in 20 ml dichloromethane in a screw-cap vial. One milliliter was then collected and evaporated to dryness under nitrogen stream. The residue was dissolved in an adequate

Volume 23 Number 3, Part I Scptcmher 1990

Blanching effect of topical hydrocortisone creams 459

u 9 e / c 4 m 2 H p L C 1.0

2.0

".0

11.0

RIA

1l.0

1l.0

(ug/cm2)

0.823· X + 0.103 r

= 0.985

Fig. 1. Correlation between radioimmunoassay and HPLC methods.

Table I. In vitro release of hydrocortisone from creams (mean ± SD) Time (min)

Hytone (tLg/cm2) Synacort (tLg/cm2)

30

60

120

240

360

36.0 ± 9.7 30.0 ± 12.7

64.0 ± 7.6 51.8 ± 7.8

107.3 ± 6.2 83.8 ± 10.7

175.3 ± 5.3 129.8 ± 10.1

227.8 ± 6.6 162.8 ± 10.6

volume of either chromatographic eluent for high-performance liquid chromatography (HPLC) analysis or distilled water for radioimmunoassay analysis. For HPLC analysis, 10 Jl.I of the sample was injected on Waters HPLC system (Waters Chromatography Div., Millipore, Milford, Mass.) equipped with Du Pont ODS 7 (250 x 4.6 mm) column (Ou Pont Co., Wilmington, Del.) and 254 nm ultraviolet detector. The hydrocortisone was eluted with 40% water--60% methanol plus 0.2% acetic acid. An Amerlex cortisol radioimmunoassay kit for the direct immunoassay of hydrocortisone in human serum, plasma, or urine was adapted for the determination of hydrocortisone in tape strips used (codeIM 2121, Arnersham Corp., Arlington Hts., 111.). HPLC and radioimmunoassay methods were compared and found to yield equivalent results (Fig. 1). Because radioimmunoassay has greater rapidity and sensitivity than HPLC, radioimmunoassay was chosen for the remainder of the study.

Skin blanching assay Methodology. For the skin blanching assay, 50 tLl (approximately 44 mg/cm-) of each formula tion was placed in large Finn chambers (Epitest Ltd.; Hyryla, Finland)

and applied topically under occlusion to 1 ern? areas of the upper part of the back of 12 subjects. The test zones were separated by at least 3 em. To obtain the extent of blanching over time (similar to area under the curve), six occlusion times, identical to those employed for the penetration test (4,6,8,16,20, and 24 hours), were chosen. Blanching assessment. The blanching assessment was performed 1 hour after removing the occlusion and the excess cream. Two independent observers performed the scoring according to a scale of 0 to 4 (0, no modification of skin color; 1; slight pallor; 2, clear pallor; 3, intense pallor; 4, maximum pallor).

In vitro release Methodology. The in vitro release characteristics of the two hydrocortisone creams were determined (6 U for each formulation) with a diffusion cell and synthetic cellulose acetate membrane." The donor chamber was filled with approximately 1gm of hydrocortisone cream and the receptor chamber with 0.05 mol/L phosphate buffer (pH 5). A 100 tLl aliquot was removed at 30, 60, 120,240, and 360 minutes. The receptor chamber was refilled with fresh phosphate buffer after each sample was removed. The samples were analyzed with the use of HPLC. The results are reported in Table I.

Journal of the American Academy of Dermatology

460 Caron et al. 04.

m

c g

/ a. c m 2

..

IS

II

'1

•

"

10

11

la

11

time ~

=--

1..

111

III

1'1

II

II

ao

at

u

u

a ..

(hours) HYTONE SYNACORT

Fig. 2. Kinetics of hydrocortisone measured in strips 2 to 10.

I.

I. I. U

g

I

c

II

2

I. I. I.

o. o.

1.0

1.0

1.0

•• 0

'.0

1.0

7.0

'.0

'.0

10.0

strip Nr - 1 - HYTONE ....... SYNACORT

Fig. 3. Distribution of hydrocortisone in Synacort cream compared with Hytone cream at 6 hours.

RESULTS The comparison of the quantities of hydrocortisone in strips 2 to 10 after application of Hytone and Synacort creams as a function of time, expressed in p-gjcm2 , is shown in Fig. 2. The distribution of hydrocortisone in strips (I through 10) is shown in Fig. 3; the quantity steadily decreases with deeper layers, as expected. The results of the blanching assay are expressed as mean score versus time. The comparison of the blanching effect observed after application of two creams as a function of time is shown in Fig. 4.

For the comparison of the two formulations, both blanching scores and quantities of hydrocortisone measured in the SC were submitted to a three-way analysis of variance. For the blanching score the formulation time interaction does not reach significance (p = 0.0732), whereas the time effect is significant (p = 0.0335). The blanching response increases, reaching a maximum at 16 hours. The formulation effect is significant (p = 0.026): Hytone cream induced more blanching than Synacort cream at all times observed. For the hydrocortisone quantities measured in the

Volume 23 Nu mber 3, Part 1

Blanching effect of topical hydrocortisone creams 461

September 1990

O. 9 C

0 I"

e

0.7 0 .11 0.11

0." 0.3

O.

4

II

II

7

I

I

10

Ii

12

13

time

I"

III

II

l'

Sill

ao

2S

aa

a3

a..

(houl"'s)

.........- HYTONE SYNACORT

===--

Fig. 4. Comparison of vasoconstriction effect observed after application of Hytone and Syn acort creams, as a function of time.

LEGEND:

PD: Vasoconstridion at 16 hours

R

E S P

PK: Skin Concentration (meg Jsq. em)

o N S E

In Vitro: meg Jsq. em. released in 6 hours N=12

PD

PK

In Vitro

Fig. 5. Correlation between pharmacodynamic (PD), pharmacokinetic (PK>, and in vitro release characteristics of two hydrocortisone creams.

se, neither formulation time interaction nor time effect is significant (p = 0.3856 and 0.4818, respectively). In other words, irrespective of the formulation, the level of hydrocortisone in the SC does not appear to change with time. This steady state probably reflects rapid achievement of equilibrium between penetration and elimination relative to the se, which remain constant because of the permanent excess formulation present on the surface throughout the test period. Nevertheless, the respective levels of hydrocortisone in the se are different (p = 0.0001); the hydrocortisone level of Hytone cream was higher than that of Synacort cream. The amount of hydrocortisone in the tape strippings indicates the relative release of active substance from

the formulations that becomes available for further penetration into the skin. DISCUSSION

Although there is no direct correlation between the results of the pharmacokinetic and pharmacodynamic measurements, both give clear evidence of a difference in drug level in the se and pharmacologic activity of the two formulations. Two explanations can be considered: • The quantities applied in the two tests were too large and different: 44 rug/ern? for the blanching test and 8 mg/cm? for the penetration test. The reasons for selecting different amounts of hydrocortisone creams were that ( 1) the blanching effect

Journal of the American Academy of Dermatology

Caron et al. of hydrocortisone is very weak under nonocclusive conditions and (2) the 1 em? surface under the large Finn Chambers is too small to obtain skin strips for analysis; a larger surface (3 cm-) was needed for this. Therefore these important quantities remained in excess during the entire penetration period. There is a discrepancy between the mechanism of the two phenomena: the penetration procedure is assumed to be a constant passive phenomenon, whereas blanching activity occurs via the cutaneous receptors (the limiting step for the appearance of the pharmacologic effect). 5 The in vitro release characteristics of the two hydrocortisone creams show a significantly different release pattern. The Hytone cream released higher amounts of hydrocortisone compared with Synacort cream. This agrees with the pharmacodynamic (vasoconstriction) and pharmacokinetic (SC hydrocortisone content) measurements. The comparison of these determinations is shown in Fig. 5. The use of synthetic membrane in deter-

mining in vitro release properties obviates the problems associated with animal and human skin membranes. The observed relationship between in vitro and in vivo measurements support the in vitro procedure. Thus these data show that a diffusion cell system with a synthetic membrane can be used as a quality control procedure to assure lot-to-lot uniformity and bioequivalence, once bioavailability has been measured. REFERENCES I. McKenzie AW, Stoughton RB. Method for comparing percutaneous absorption of steroids. Arch Dermatol 1962; 86:608-10. 2. Cornell RC, Stoughton RB. Correlation of the vasoconstriction assay and clinical activity in psoriasis. Arch Dcrmatol 1985;121 :63-7. 3. Dupuis 0, Rougier A. In vivo relationship between horny layer and resultant effect and percutaneous absorption in human and rat. J Invest Dermatol 1984;82:353-6. 4. Shah VP, Elkins J, Lam SY, Skelly JP. Determination of in vitro drug release from hydrocortisone creams. lnt J Pharm 1989;53:53-9. 5. Marks R, Barlow JW, Funder JW. Steroid-induced vasoconstriction: glucocorticoid antagonist studies. J Clin Endocrinol Metab 1982;54: 1075-7.

Treatment of hirsutism with the pure antiandrogen flutamide Leonello Cusan, MD, Phl)," Andre Dupont, MD, Phl)," Alain Belanger, Phl)," Roland R. Tremblay, MD, PhD,b Gilles Manhes, MD,a and Fernand Labrie, MD, PhDa

Ste. -Foy, Quebec, Canada The effectiveness of the antiandrogen f1utamide in combination with an oral contraceptivewas studied in 20 patients with moderate to severehirsutism.Eight patients had no previous therapy, whereas 12 had failed to respond to oral contraceptives, spironolactone, or dexamethasone therapy. Treatment with the antiandrogen flutamide (250 mg twicedaily) and an oral contraceptive (Ortho 1/35) resulted in a particularly rapid and marked decrease in the total hirsutism score, which reached the normal range at 7 months. Seborrhea, acne,and hair loss score were also rapidly corrected. Treatment was associated with a decrease in plasma luteinizing hormone, progesterone, and estradiol levels. Plasma sex hormone-binding globulin levels were initially low in 18 of 20 patients but increasedsignificantly during therapy. No clinically significant side effects were observed. (1 AM ACAD DERMATOL1990;23:4629.) From the Laboratory of Molceular Endocrinology" and the Laboratory of Hormonal Biorcgulatlon,"Centre Hospitalicr de l'Universite Laval Research Center. Accepted for publication Dec. 18,1989. Reprint requests: Fernand Labrie, MD, PhD, Laboratory of Molecular Endocrinology, CnUL Research Center, 2705 Laurier Blvd., SteFay, Quebec GIV 402, Canada.

16/1/19176

462

Hirsutism and related signs such as acne, alopecia, and menstrual irregularities are androgen dependent and respond to antiandrogen therapy.'? Among the drugs with antiandrogen properties, only cyproterone acetate (CPA) and spironolactone have been used in clinical trials. 2,4-6 The efficacy of these two drugs has been limited by their relatively weak