Bioidentical Compounded Hormones: a Pharmacokinetic Evaluation

Home , Compounding

Maturitas 74 (2013) 375–382

Contents lists available at SciVerse ScienceDirect

Maturitas

j ournal homepage: www.elsevier.com/locate/maturitas

Bioidentical compounded hormones: A pharmacokinetic evaluation

ଝ

in a randomized clinical trial

a,∗ b c d e

Richa Sood , Roger A. Warndahl , Darrell R. Schroeder , Ravinder J. Singh , Deborah J. Rhodes ,

f g a

Dietlind Wahner-Roedler , Rebecca S. Bahn , Lynne T. Shuster

Women’s Health Clinic, Division of General Internal Medicine, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905, USA

Division of Pharmacy, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905, USA

Health Sciences Research, Division of Biomedical Statistics and Informatics, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905, USA

Department of Laboratory Medicine and Pathology, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905, USA

Executive Health, Division of Preventive and Occupational Medicine, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905, USA

Breast Clinic, Division of General Internal Medicine, Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905, USA

Division of Endocrinology and Metabolism; Mayo Clinic, 200 First Street SW, Rochester, Minnesota 55905, USA

a r t i c l e i n f o a b s t r a c t

Article history: Objective: Bioidentical compounded hormone therapy is popular among patients, but providers do not

Received 10 December 2012

have pharmacokinetic information or dosing guidelines for these preparations. Our objective was to com-

Received in revised form 11 January 2013

pare the pharmacokinetics of the commonly used compounded preparations with conventional hormonal

Accepted 12 January 2013

preparations that are considered bioequivalent in practice.

Methods: We conducted a randomized, blinded, four-arm 16-day clinical trial of forty postmenopausal

women assigned to one of three doses of a compounded estrogen cream (Bi-est (80:20); 2.0, 2.5, or

Keywords: TM

3.0 mg) + compounded oral progesterone 100 mg, or to a conventional estradiol patch (Vivelle-Dot

Bioidentical TM

Compounded 0.05 mg) + Prometrium 100 mg. Serum levels of estrone, estradiol, estriol, and progesterone were

obtained at multiple time intervals during the ﬁrst 24-h, and at steady-state.

Hormone therapy

Pharmacokinetics Results: Results were analyzable for 37/40 women. Study medications were well tolerated. The AUC at

24 h and at steady-state for estrogens remained consistently lower for all doses of Bi-est tested relative

to the patch. The difference was statistically signiﬁcant for Bi-est 2.0 mg (AUC-estradiol = 181 vs. 956;

p < 0.001) and 2.5 mg (AUC-estradiol = 286 vs. 917; p < 0.001). Estriol levels remained low in all study

arms. Serum progesterone levels were comparable in conventional vs. compounded groups.

Conclusions: This pharmacokinetic trial showed that the currently used doses of compounded hormones

yield lower levels of estrogen compared to the standard-dose estradiol patch. To ﬁnd comparable doses,

further studies are needed. This successfully conducted randomized controlled study attests to the fea-

sibility of using a similar design in the setting of a larger clinical trial.

1. Introduction proportion of these women also experience vaginal dryness [3], uri-

nary incontinence [3], decline in sexual interest and satisfaction [4],

With an aging population globally, the world’s total menopausal mood ﬂuctuations [5,6], sleep disturbances [7–9], and changes in

population, estimated at 476 million in 1990, is expected to rise to memory and cognition, for which hormone therapy is often sought.

1.2 billion by 2030 [1]. In the United States, approximately 75%–80% The collective impact of these symptoms on women’s well-being is

of postmenopausal women report vasomotor symptoms [2]. A large enormous, and the need for safe treatments is compelling [10].

Conventional hormone therapy consisting of Food and Drug

Administration (FDA)-approved products is the standard of care

in the United States when treatment is indicated for menopausal

Abbreviations: CRU, Clinical Research Unit; CVs, coefﬁcient of variations.

ଝ symptom relief. However, randomized trials such as the Heart and

Trial registration: linicaltrials.gov identiﬁer: NCT00864214.

∗

Estrogen/Progestin Replacement Study [11] and Women’s Health

Corresponding author at: Women’s Health Clinic, Mayo Clinic, 200 First Street

SW, Rochester, Minnesota 55905, USA. Tel.: +1 507 538 7103; fax: +1 507 266 3988. Initiative [12] raised concerns about the safety of conventional

E-mail addresses: [email protected] (R. Sood), [email protected] TM TM

hormones like Premarin and Provera , leading to a marked

(R.A. Warndahl), [email protected] (D.R. Schroeder),

reduction in their use and creating a demand for safer alternatives.

[email protected] (R.J. Singh), [email protected] (D.J. Rhodes),

Bioidentical hormones have gained popularity as possible

[email protected] (D. Wahner-Roedler), [email protected]

(R.S. Bahn), [email protected] (L.T. Shuster). alternatives to conventional hormones. The term “bioidentical”

376 R. Sood et al. / Maturitas 74 (2013) 375–382

Table 1

Study design.

No. of patients/arm Randomization

Active intervention Placebo Intervention phase (16 days)

Estrogen Progesterone Day 1 Days 15 and 16

10 Bi-est 2.0 mg Compounded Patch Serial measurements Serial measurements

progesterone 100 mg E1, E2, E3 E1, E2, E3;

10 Bi-est 2.5 mg Compounded Patch Baseline and 24 h Single

progesterone 100 mg progesterone level end-of-treatment

progesterone level

10 Bi-est 3.0 mg Compounded Patch Monitoring for adverse Monitoring for adverse

progesterone 100 mg effects effects

10 Estradiol 0.05 mg patch Micronized Cream

(Vivelle-Dot ) progesterone 100 mg capsule

(PrometriumTM)

implies a chemical and molecular structure precisely the same as limit of normal) and a negative mammogram within the last 11

its endogenous human hormone counterpart. Structural similarity months. Presence or absence of menopausal symptoms was not

by some is believed to correlate with safety, a perception that has used as an eligibility criterion. Subjects were excluded using the

led to a sharp increase in the popularity of bioidentical hormones. following criteria: estrogen levels >35 pg/mL; >10 years from last

Currently there are multiple FDA-approved conventional hor- menstrual period; a history of cancer of the breast, uterus or ovary,

mone products that are plant-derived and bioidentical in chemical coronary artery disease, stroke, dementia, migraine, deep venous

structure [13]. However, likely due to extensive media coverage, thromboembolism, active liver or gall bladder disease, uncontrolled

requests from women for bioidentical hormones are predomi- hypertension, diabetes, or lupus; currently smoking; alcohol or

nantly for the compounded preparations. In the absence of adequate substance abuse; family history of premenopausal breast or ovarian

research, no clear guidelines exist for choosing the most appropri- cancer; or postmenopausal breast cancer in ≥2 relatives. Recent use

ate type, dose, or regimen of bioidentical compounded hormones. of hormone therapy was allowed with an adequate washout period

Lacking this data, it is challenging to study these preparations for (at least 1 week for vaginal hormones, 4 weeks for transdermal

safety and efﬁcacy. hormones, and 8 weeks for oral hormones). Additional exclusions

To the best of our knowledge, no pharmacokinetic studies have included peanut allergy, current use of isoﬂavone-containing prod-

compared conventional with compounded bioidentical hormones. ucts, and drugs or herbs that might affect metabolizing enzymes.

Compounding pharmacists generally consider Bi-est (referring to

estriol plus estradiol) in a dose of 2.5 mg (for example, in a ratio of

2.2. Study treatment

80% estriol plus 20% estradiol, which consists of 2 mg estriol and

0.5 mg estradiol) as approximately equivalent to a mid-range dose

Eligible women were randomized to one of four treat-

of an estradiol-containing patch such as Vivelle-Dot 0.05 mg.

ment arms in a double-blinded fashion. Three study arms

Thus, using Bi-est with 80% estriol and 20% estradiol, and comparing

included compounded estradiol–estriol cream (Bi-est) in vary-

the calculated estradiol equivalence of the compounded prepa-

ing dosage, along with a placebo skin patch, and compounded

rations with the daily referenced estradiol dose of the patch, we

oral micronized progesterone capsules. The fourth arm included

assigned the three strengths of creams to create a reasonable range TM

an estradiol-containing patch (Vivelle-Dot ), placebo cream, and

for comparison.

commercially available oral micronized progesterone capsules

In order to lay the groundwork for future trials, we conducted a

(Prometrium) (Table 1).

pharmacokinetic study as a randomized, blinded clinical trial. Our

Bi-est (80:20) 2.5 mg was chosen as it is considered equivalent

intent was to compare estrogen and progesterone levels obtained

to 0.05 mg estradiol patches in compounding practice. Bi-est was

following the administration of bioidentical compounded hormone

compounded in Vanicream and dispensed in premarked individ-

therapy preparations both with themselves and with conventional

ual syringes by an experienced compounding pharmacist (RAW),

hormone therapy.

who was blinded to the participants’ study group assignment.

Women in study arm one received Bi-est 2.0 mg (1.6 mg estriol and

2. Methods 0.4 mg estradiol), in arm two Bi-est 2.5 mg (2 mg estriol and 0.5 mg

estradiol), and arm three Bi-est 3.0 mg (2.4 mg estriol and 0.6 mg

2.1. Subjects estradiol). Women in study arm four received Vivelle-Dot 0.05 mg

patches.

The study was approved by the Mayo Clinic Institutional Review On day one of the study, participants were admitted to the hos-

Board. Since estriol is not an approved product in the United States, pital in the CRU, where they were instructed to apply their ﬁrst dose

an IND (investigational new drug) application was ﬁled with the under supervision by trained nurses. The creams were applied to

US-FDA. This was a Phase I, blinded, randomized, four-arm study. a speciﬁed area of the forearm by gentle rubbing for 1 min. The

Subjects were recruited from Rochester, MN, and surrounding areas Vivelle-Dot patch was applied to the skin of the lower abdomen

through advertisements and news media releases. Eligibility was following the manufacturer’s directions. The capsules were admin-

based on the following criteria:women 40–60 years old, naturally istered orally under supervision. Throughout the next 24 h, serial

postmenopausal (absence of periods for ≥1 year or amenorrhea blood samples were obtained at speciﬁed intervals to measure the

for ≥6 months and FSH ≥40 IU/L) or surgically postmenopausal serum levels of estrogen fractions and progesterone (Table 2). After

(menopause induced by removal of ovaries), able to understand 24 h of the CRU stay, participants were discharged and continued

and sign an informed consent, and willing to stay overnight in a taking study medications as instructed. They returned on the 15th

Clinical Research Unit (CRU). Eligible women had normal results on day of the study and were admitted for 12 h during which time

screening tests (AST, creatinine, and TSH within 20% of the upper serial blood samples were obtained to measure the steady-state

R. Sood et al. / Maturitas 74 (2013) 375–382 377

Table 2

Study intervention schedule.

Intervention phase

Venipuncture Day 1 (hours after initial application of Days 7 Venipuncture Days 15 and16 (hours after ﬁnal application

cream/patch) and 13 of cream/patch)

Hours −1 2 5 8 12 18 24 −1 2 5 8 12 24 30

E1, E2, E3 x x x x x x x x x x x x x x

Progesterone x x x Study

Coordinator x

Phone call

levels of estrogen and progesterone. Participants returned on day was used to compare each of the Bi-est groups to the Vivelle-

16 for the last set of blood draws. They returned any remaining sup- Dot patch group. As a secondary analysis, progesterone levels

ply of medications. Compliance was assessed by counting syringes, were summarized and compared between the Bi-est arms and

patches, and capsules that were returned, as well as by self report. Vivelle-Dot group using the rank sum test. In all cases, two-tailed

p-values ≤ 0.05 were considered statistically signiﬁcant. The fre-

quency of adverse events was summarized according to treatment

2.3. Hormone measurements

group and compared across groups using the Fisher exact test.

Estrogen fractions (Estrone E1, E2) were measured by liquid

chromatography–tandem mass spectrometry (ThermoFisher Sci- 3. Results

entiﬁc, Franklin, MA and Applied Biosystems-MDS Sciex, Foster

City, CA). For E1, the intra-assay coefﬁcient of variations (CVs) were 3.1. Participants

10.9%, 3.4%, 5.1%, and 3.5% at 6.7, 29, 109, and 332 pg/mL respec-

tively, and the inter-assay CVs were 8.1%, 6.9%, 5.2%, 6.3%, and 6.7% Fifty-two postmenopausal women were screened; 12 of those

at 6.4, 26, 58, 120, and 336 pg/mL respectively. For E2, the intra- were excluded and all 40 women who were randomized into the

assay CVs were 12.4%, 3.1%, 5.0%, and 3.5% at 7.2, 29, 109, and trial completed the study (Fig. 1). The average age of the study par-

325 pg/mL respectively, and the inter-assay CVs were 9.2%, 8.6%, ticipants was 54 years; their average BMI was 29 kg/m (Table 3).

9.0%, 6.6%, and 4.8% at 7.0, 24, 61, 125, and 360 pg/mL respec- The compliance rate for use of Bi-est creams and Vivelle-Dot patch

tively. The limit of quantitation for both E1 and E2 was 2.5 pg/mL. was 100%. Only one participant returned a single unused proges-

Values less than 2.5 pg/mL were reported as <2.5 pg/mL. Estriol terone capsule. Two participants whose initial screening estradiol

(E3) was measured by using the Access Unconjugated Estriol levels were in the postmenopausal range had elevated baseline

assay, which was a competitive binding immunoenzymatic assay estradiol levels on repeat testing before administration of the

measured on the DxI automated immunoassay system (Beckman study medications (82 pg/mL and 151 pg/mL respectively) and were

Instruments, Chaska, MN). The inter-assay CVs were 6.4%, 4.6%, excluded. A third participant with non-physiologic estradiol lev-

and 4.9% at 1.04, 2.78, and 5.65 ng/mL respectively, and the lowest els (1500 pg/mL), presumably from contamination with estrogen

reportable result was 0.07 ng/mL. Progesterone was measured by a cream at the venipuncture site, was excluded after the study medi-

competitive binding immunoenzymatic assay on the DxI 800 auto- cations were administered. Data were analyzable for 37/40 women.

mated immunoassay system (Beckman Instruments, Chaska, MN).

The inter-assay CVs were 16.5%, 10.1%, and 7.7% at 0.76, 9.2, and 3.2. Estrogen pharmacokinetics

29.6 ng/mL respectively. The lower limit of the clinically reportable

range for progesterone was 0.08 ng/mL. Average concentrations and AUC were calculated for estrogen

fractions after initial dosing and at steady-state (Tables 4 and 5). The

Bi-est cream groups showed a slight trend toward increasing serum

2.4. Statistical analysis

levels of estrogen with an increasing dose of estrogen. However,

participants had wide ﬂuctuations in their estradiol level with Bi-

The concentration of each estrogen fraction at the start of the

est, both after the initial administration as well as at steady-state

study period (C0) and at 24 h after treatment application (C24)

(Figs. 2 and 3). In contrast, the estrogen absorption with Vivelle-

was measured, as was the difference between the baseline and

Dot patches was more consistent across the ten participants given

24-h concentrations (C24 − C0). We also calculated the average con-

the patch, and the pattern was similar to what is reported in the

centration and area under the curve (AUC) for each of the three

literature (Fig. 2).

estrogen fractions over the ﬁrst 24 h. Due to the marked variabil-

ity observed in the individual concentration–time proﬁle plots,

maximum concentration (Cmax) and time to maximum concen- 3.2.1. Estradiol

tration (Tmax) were not analyzable. The results from each of the In the Bi-est groups, for the majority of subjects a peak estradiol

three cream groups were compared with the Vivelle-Dot patch level was absent after a single-dose administration (Fig. 2). The 24-h

group. Progesterone levels were measured at baseline, 24 h, and E2-AUC was smaller in all Bi-est groups compared to the estradiol

at the end of study. We calculated the difference between the patch group, with the difference being statistically signiﬁcant for

mean concentration at 24 h and at the end of the study com- Bi-est 2.0 and 2.5 mg. The steady-state E2-AUC (Days 15, 16) was

pared to baseline. The combined results from the compounded also smaller for all Bi-est groups compared to the patch, with the

micronized progesterone groups were compared against the con- difference being statistically signiﬁcant for the Bi-est 2.0 and 2.5 mg

ventional Prometrium group. groups (Table 4).

Subject characteristics were summarized using

mean standard deviation, median (min, max) for continu- 3.2.2. Estrone

ous variables, and frequency counts for categorical variables. E1, Estrone (E1) levels in the Bi-est groups remained lower than

E2, and E3 parameters described previously were summarized for the patch, both after initial administration of the medica-

separately for each of the 4 treatment groups. The rank sum test tion, as well as at steady-state (Fig. 4). The 24-h E1-AUC and

378 R. Sood et al. / Maturitas 74 (2013) 375–382

Assessed for eligibility

n = 52

Exclu ded (n = 12)

Enrollment Not meetin g inclusion

criteria (n = 11)

Decli ned to participate

(n = 1)

Other reasons (n = 0)

Randomized (n = 40)

Bi-est 2.0 mg Bi-est 2.5 mg Bi-est 3.0 mg Vivelle-Dot

Allocation (Arm 1; n = 10) (Arm 2; n = 10) (Arm 3; n = 10) 0.05 mg patch

(Arm 4; n = 10)

Foll ow-up Discontinued Discontinued intervention Discontinued Discontinued

interve ntion (n=3) intervention intervention

(n =0) (2 participa nts excluded (n =0) (n =0)

before study drug

administrati on due to elevated

repeat bas eline estradiol;

1 participant excluded after

starting the study medication

due to supra-physiologic

estrogen levels possibly

secondar y to contamination)

Analysis Analyzed Analyzed Analyzed Analyzed

(n = 10) (n = 7) (n = 10) (n = 10)

Fig. 1. Flow diagram.

steady-state E1-AUC were smaller across all Bi-est groups com- due to very low serum values, average E3 values were comparable

pared to the patch, and the difference was statistically signiﬁcant at baseline. The 24-h median AUC for the Bi-est 2.0 mg arm and the

for the Bi-est 2.0 and 2.5 mg arms (Table 5). steady-state median AUC for the Bi-est 2.0 and 2.5 mg arms were

signiﬁcantly lower than the patch (p < 0.05) (Fig. 5).

3.2.3. Estriol

Most participants had E3 levels below the lower limit of detec- 3.3. Progesterone pharmacokinetics

tion of the assay, both at baseline and at steady-state. All E3 assays

were run in triplicate to evaluate the reliability of these results. Progesterone levels were analyzed in the 37/40 women whose

Acknowledging the limitation of a wide coefﬁcient of correlation estrogen levels were conﬁrmed to be menopausal. The levels at

Table 3

Baseline characteristics of study participants.

Characteristic Bi-est 2.0 mg Bi-est 2.5 mg Bi-est 3.0 mg Vivelle-Dot

(n = 10) (n = 7) (n = 10) 0.05 mg (n = 10)

Age, years

Mean SD 53.7 ± 4.3 55.1 ± 3.6 52.9 ± (3.8) 53.8 ± 2.7

Median (min, max) 54 (44, 59) 55 (50, 59) 53 (48, 59) 54 (48, 57)

BMI, kg/m

Mean ± SD 29.9 ± 5.1 28.8 ± 9.3 29.3 ± 6.5 28.1 ± 5.8

Median (min, max) 30 (23, 38) 27 (18, 42) 28 (21, 44) 28 (20, 38)

Prior conventional hormone therapy use

No 6 3 2 2

Yes 4 4 8 8

Prior compounded hormone therapy use

No 10 7 10 9

Yes 0 0 0 1

R. Sood et al. / Maturitas 74 (2013) 375–382 379

Table 4

Estradiol pharmacokinetics at baseline, 24 h and steady-state in a phase I randomized clinical trial for evaluating estrogen pharmacokinetics.

Characteristic Bi-est 2.0 mg Bi-est 2.5 mg (n = 7) Bi-est 3.0 mg Vivelle-Dot

(n = 10) (n = 10) 0.05 mg (n = 10)

BASELINE – single dose (all results measured in pg/mL)

Mean ± SD 5.7 ± 1.1 6.5 ± 2.3 10.0 ± 9.2 8.5 ± 5.8

Median (min, max) 5.0 (4.9, 7.7) 5.2 (4.9, 11.0) 5.1 (4.9, 34.0) 6.0 (4.9, 23.0)

C24

± a a b

Mean SD 8.7 3.8 9.0 ± 3.1 51.0 ± 90.7 57.5 ± 28.5

Median (min, max) 8.2 (4.9, 17.0) 10.0 (4.9, 12.0) 26.0 (4.9, 306.0) 54.5 (19.0, 110.0)

C24 − C0

± a a c

Mean SD 3.0 ± 3.3 2.5 ± 2.9 50.0 ± 91.9 49.0 ± 25.9

Median (min, max) 1.6 (0.0, 9.4) 1.0 (0.0, 6.8) 13.6 (0.0, 301.1) 44.4 (14.1, 89.1) Average

a a

± ± ± ±

Mean SD 7.7 2.6 7.3 2.5 26.5 27.7 35.8 19.4

Median (min, max) 7.0 (5.1, 13.1) 6.5 (4.9, 11.9) 16.0 (5.1, 96.4) 33.2 (10.8, 71.4) AUCd

a a

Mean ± SD 193 ± 76 177 ± 62 730 ± 840 1043 ± 562

Median (min, max) 181 (124, 381) 153 (118, 293) 405 (122, 2919) 956 (306, 2012)

End of study – steady-state (all results measured in pg/mL)

a b b

Mean ± SD 9.5 ± 3.2 25.2 ± 34.9 20.9 ± 14.1 31.7 ± 21.5

Median (min, max) 9.4 (5.5, 16.0) 11.0 (4.9, 103.0) 19.0 (5.4, 52.0) 26.0 (13.0, 91.0)

C24

a a c

Mean ± SD 12.3 ± 8.9 12.7 ± 7.9 22.2 ± 19.0 62.7 ± 39.9

Median (min, max) 9.8 (4.9, 36.0) 11.0 (6.9, 29.0) 14.5 (5.5, 70.0) 51.5 (18.0, 129.0)

C24 − C0

c c c

Mean ± SD 2.8 ± 8.7 −12.6 ± 36.6 1.4 ± 11.9 31.0 ± 30.4

Median (min, max) 0.0 (−5.5, 26.0) −0.6 (−92.0, 19.7) 0 (−26.0, 18.0) 26.0 (2.0, 98.0) Average

± a b

± ±

Mean SD 11.8 4.9 17.5 10.7 32.2 32.4 44.8 26.8

Median (min, max) 10.8 (6.3, 24.3) 12.4 (5.8, 33.8) 21.7 (10.4, 117.3) 33.0 (12.1, 95.2) AUCd

a c

Mean ± SD 286 ± 118 374 ± 199 728 ± 643 1247 ± 760

Median (min, max) 269 (149, 592) 337 (143, 641) 540 (262, 2340) 917 (321, 2550)

Rank sum test p < 0.001 compared to Vivelle-Dot 0.05 mg.

Rank sum test p < 0.05 compared to Vivelle-Dot 0.05 mg.

Rank sum test p < 0.01 compared to Vivelle-Dot 0.05 mg.

At baseline (after a single dose) estrogen fractions were measured at 0, 2, 5, 8, 12, 18, and 24 h. At the end of study (steady-state) estrogen fractions were measured at 0,

2, 5, 8, 12 and 24 h. The area under the curve (AUC) was calculated according to the trapezoidal rule.

baseline were comparable in all 4 groups. The levels increased com- coefﬁcient of correlation between BMI and E3 was noted to be −0.33

parably at 24 h and at steady-state with administration of both the at steady-state, and the p value was not statistically signiﬁcant

compounded and conventional bioidentical progesterone prepara- (Fig. 6).

tions; no intergroup differences were noted (Table 6).

3.5. Safety

3.4. Effect of BMI on estrogen

The study medications were well tolerated. No patient dis-

There was wide variability in the BMI of the study partici- continued the study due to medication side effects. The majority

pants. Since the enzymes in body fat are involved in estrogen (57%) of reported symptoms were deemed to be unrelated to study

metabolism, we plotted E3 levels against BMI. We used the Vivelle-

Dot patch group, as this group had no exogenous E3 administered

Steady State E2

and thus any rise in E3 was assumed to be from interconversion. The

90 Biest 2.0 mg

Biest 2.5 mg

Day 1 E2 80

Biest 3.0 mg Vivell e-Dot 90 Biest 2.0 mg 70 Biest 2.5 mg 80 Biest 3.0 mg 60 70 Vivelle-Dot 50 60 50 40 40 30

30 20 20 10 10

0 5 10 15 20 25 30 0 51015202530

Hours Hours

Fig. 2. Mean estradiol across groups – initial 24-h curves. Fig. 3. Mean estradiol across groups – steady-state (Days 15/16) curves.

380 R. Sood et al. / Maturitas 74 (2013) 375–382

Table 5

Estrone pharmacokinetics at baseline, 24 h and steady-state in a phase I randomized clinical trial for evaluating estrogen pharmacokinetics.

Characteristic Bi-est 2.0 mg Bi-est 2.5 mg (n = 7) Bi-est 3.0 mg Vivelle-Dot

(n = 10) (n = 10) 0.05 mg (n = 10)

BASELINE – single dose (all results measured in pg/mL)

Mean ± SD 23.9 ± 5.7 24.6 ± 7.8 31.3 ± 16.8 29.2 ± 10.3

Median (min, max) 24 (26, 35) 23 (24, 37) 27.5 (12, 54) 26.5 (17, 45)

C24

a a

Mean ± SD 30.4 ± 10.0 29.7 ± 6.8 39.8 ± 18.4 51.3 ± 16.3

Median (min, max) 28.5 (18, 52) 34 (21, 36) 40.5 (15, 67) 51.5 (27, 89)

C24 − C0

a b a

Mean ± SD 6.5 ± 8.8 5.1 ± 4.7 8.5 ± 8.4 22.1 ± 11.6

Median (min, max) 2.5 (−2, 27) 4.0 (−3, 11) 8.0 (−1, 27) 22 (10, 47) Average

c c

Mean ± SD 23.6 ± 6.7 22.5 ± 8.6 31.1 ± 16.6 35.1 ± 10.4

Median (min, max) 22.9 (14.7, 33.6) 20.7 (12.3, 35.4) 27.2 (12.0, 58.6) 32.7 (20.7, 53.0) AUCd

a c

Mean ± SD 563 ± 168 529.4 ± 219.5 747 ± 404 876 ± 254

Median (min, max) 550 (346, 830) 495 (255, 854) 656 (283, 1427) 844 (510, 1339)

End of study – steady-state (all results measured in pg/mL)

a c

Mean ± SD 28.6 ± 7.4 26.6 ± 9.4 40.7 ± 21.1 50.7 ± 29.6

Median (min, max) 27.5 (19, 42) 29 (11, 37) 36 (15, 97) 42.5 (18, 127)

C24

b b c

Mean ± SD 30.6 ± 8.7 28.6 ± 7.0 38.0 ± 16.4 61.7 ± 31.8

Median (min, max) 28 (21, 46) 27 (19, 38) 35.5 (16, 59) 56 (31, 146)

C24 − C0

Mean SD 2.0 ± 5.1 2.0 ± 8.1 −2.7 ± 9.0 11.0 ± 13.8

Median (min, max) 3 (−5, 10) 1 (−11, 15) −0.5 (−23, 5) 13.5 (−13, 29) Average

b a

Mean ± SD 26.2 ± 5.8 25.6 ± 8.7 36.1 ± 17.0 48.8 ± 23.0

Median (min, max) 25.8 (19.5, 34.8) 26.5 (11.5, 40.2) 31.8 (14.0, 62.7) 42.4 (20.3, 107.2) AUCd

b a

Mean ± SD 623 ± 137 615 ± 197 836 ± 376 1209 ± 556

Median (min, max) 625 (476, 854) 605 (282, 934) 757 (334, 1437) 1072 (514, 2613)

Rank sum test p < 0.01 compared to Vivelle-Dot 0.05 mg.

Rank sum test p < 0.001 compared to Vivelle-Dot 0.05 mg.

Rank sum test p < 0.05 compared to Vivelle-Dot 0.05 mg.

At baseline (after a single dose) estrogen fractions were measured at 0, 2, 5, 8, 12, 18, and 24 h. At the end of study (steady-state) estrogen fractions were measured at 0,

2, 5, 8, 12 and 24 h. The area under the curve (AUC) was calculated according to the trapezoidal rule.

medications (e.g. upper respiratory infection, gastroenteritis). Of 4. Discussion

the remaining symptoms, most were mild to moderate, and most

(71%) resolved by day 23 at the time of ﬁnal follow-up phone Our study was designed as a comparative prospective pharma-

visit. The reported symptoms included abdominal bloating and pain cokinetic trial, with the goal to compare the pharmacokinetics of

(n = 6, 15%), upper respiratory infection (n = 5, 12.5%), ﬂuid reten- the conventional and compounded hormonal preparations that are

tion (n = 3, 7.5%), gastroenteritis (n = 3, 7.5%), skin irritation (n = 3, considered bioequivalent in clinical practice. To the best of our

7.5%), progesterone withdrawal bleed (n = 2, 5%), and insomnia, knowledge, this is the ﬁrst study of its kind in the literature.

dizziness, fatigue and vaginal discharge in one patient each (n = 1, Contrary to our expectations, we found that the dose of Bi-est

2.5% each). 2.5 mg, commonly prescribed in lieu of the estradiol 0.05 mg patch,

Table 6

Progesterone pharmacokinetics at baseline, 24 h and steady-state in a phase I randomized clinical trial for evaluating estrogen pharmacokinetics.

Characteristic Bi-est 2.0 mg Bi-est 2.5 mg (n = 7) Bi-est 3.0 mg Vivelle-Dot

(n = 10) (n = 10) 0.05 mg (n = 10)

Mean ± SD 0.38 ± 0.23 0.60 ± 0.52 0.40 ± 0.27 0.65 ± 0.35

Median (min, max) 0.30 (0.09, 0.80) 0.63 (0.09, 1.59) 0.32 (0.09, 0.86) 0.52 (0.29, 1.37)

C24

Mean SD 1.82 ± 1.85 1.72 ± 0.72 1.71 ± 0.62 1.44 ± 0.50

Median (min, max) 1.29 (0.56, 6.89) 1.73 (0.31, 2.52) 1.70 (0.80, 2.59) 1.38 (0.92, 2.54)

CSteady-state

Mean ± SD 0.88 ± 0.37 1.01 ± 0.57 0.95 ± 0.40 1.16 ± 0.53

Median (min, max) 0.90 (0.22, 1.27) 1.12 (0.23, 1.87) 0.95 (0.44, 1.51) 1.18 (0.18, 2.16)

C24 − C0

Mean SD 1.43 ± 1.92 1.12 ± 1.07 1.31 ± 0.62 0.80 ± 0.69

Median (min, max) 0.92 (−0.04, 6.74) 1.00 (−0.49, 2.43) 1.44 (0.48, 2.28) 0.68 (−0.44, 2.08)

Css – C0

Mean SD 0.50 ± 0.47 0.41 ± 0.93 0.55 ± 0.39 0.52 ± 0.74

Median (min, max) 0.59 (−0.58, 1.09) 0.49 (−1.01, 1.68) 0.60 (−0.02, 1.20) 0.58 (−1.19, 1.24)

Note: All results measured in ng/mL.

R. Sood et al. / Maturitas 74 (2013) 375–382 381

Steady State E1 Steady_state E3 (24 hours) vs BMI

0.08

80 Biest 2.0 mg

Biest 2.5 mg 0.07 r = -0.33

70 Biest 3.0 mg p=0.35 Vivell e-Dot 0.06 60 0.05 50 0.04 40 0.03

30 Steady State E3 0.02 20 0.01 10 0 18 20 22 24 26 28 30 32 34 36 38 40

0 0 51015202530 BMI Hours

Steady_state E2 (24 hours) vs BMI

Fig. 4. Steady state E1. 140

120 r = -0.21

yielded much lower estrogen levels compared to the patch. Even p=0.56

100

a higher dose of Bi-est 3.0 mg yielded lower estrogen levels than

the 0.05 mg Vivelle-Dot patch, although the difference didn’t reach 80 statistical signiﬁcance. 60

Steady State E2

We found that patterns of estrogen absorption with Bi-est

creams were highly variable and showed no consistent peak of

absorption. Vivelle-Dot, on the other hand, showed a pattern

of absorption similar to what is reported in the literature [14]. 0

18 20 22 24 26 28 30 32 34 36 38 40

These differences may be explained by the interindividual vari-

BMI

ations in skin physiology, blood ﬂow, and hormone metabolism.

More importantly, use of different vehicles of administration

may also explain these differences, as seen from studies of Steady_state E1 (24 hours) vs BMI

160

estrogen in hydroalcoholic gels [14,15]. Our results cannot be

compared directly to another cream-based preparation owing to 140 r = -0.32

p=0.37

non-availability of any such FDA-approved hormone cream. 120

FDA-approved E3 assays test the high E3 levels of pregnancy. 100

Their reliability in detecting the extremely low postmenopausal 80

E3 levels is unproven. Acknowledging this limitation, our results

showed a small rise in E3 in all treatment groups, regardless of Steady State E1

whether E3 was administered exogenously or not. We observed

a surprising trend toward increased E3 levels in the E2 patch

group compared to the E2+E3 cream groups, which may be partly 18 20 22 24 26 28 30 32 34 36 38 40

explained by endogenous conversion from higher E2 in the patch BMI

group to higher E3 levels. Overall, we observed only a small increase

in E3 and hypothesize the possibility of a rapid conversion of E3 Fig. 6. Relationship of steady-state E1, E2, and E3 levels with body mass index (BMI).

to some of its metabolites, or interference in its absorption due to

Vanicream. Thus, future studies testing E3 metabolites and utilizing

other vehicles of absorption are needed. In our study, progesterone derived from a placebo effect. Our ﬁndings also raise the question

levels were comparable at several time points after initial dosing as to whether the observed estrogen levels with the compounded

as well as at steady-state in all groups, thereby suggesting that formulations in current doses are sufﬁcient to potentiate any bone

oral absorption of natural micronized progesterone is comparable beneﬁts in menopausal women, which needs to be studied in future

across compounded and conventional preparations. clinical trials.

From a clinical perspective, the ﬁnding of small increments in There are certain limitations to our study. First, our choice of par-

estradiol levels with low-dose Bi-est formulations raise the ques- ticular formulations and dosing was empiric because there were

tion of how much symptomatic beneﬁt attributed to these doses is no evidence-based guidelines available to direct us. Second, our

results can be generalizable to Vanicream-based compounded for-

mulations only. We chose Vanicream as the compounding medium

as it is commonly used in the compounding pharmacy at our insti-

tution. Third, although we measured E3 levels in triplicate using

a state-of-the-art automated immune assay, the test has signiﬁ-

cant variability and poor reliability at very low postmenopausal

levels. Lastly, since we tested multiple medication doses, we chose

the multiple arm randomized controlled design without cross-over

due to time and cost issues.

Our study has several strengths. First, this is the ﬁrst study

of its kind to compare the pharmacokinetics of compounded and

conventional bioidentical hormones. Second, the study design

was randomized, controlled, and blinded, which optimizes the

reliability of results. All study medications were prepared and dis-

Fig. 5. Steady state E3. pensed by the same experienced compounding pharmacist (RAW),

382 R. Sood et al. / Maturitas 74 (2013) 375–382

thereby minimizing variation in compounding. The study uti- Role of funding source

lized a strict protocol for medication application, thus reducing

the potential for variability in results attributable to inconsistent Solvay was not involved with the study design, data collection

administration of the product. Finally, the successful design and and analysis, or writing of the paper.

conduct of this study supports the feasibility of conducting research

comparing compounded hormone therapies with well-studied, Ethical approval

gold-standard hormone therapy products. This is important as

the lay public, as well as medical providers, are increasingly The study was approved by the Mayo Clinic institutional Review

seeking reliable information regarding these widely used prod- Board.

ucts.

In conclusion, this pharmacokinetic trial provides information References

regarding bioidentical compounded preparations and bioidentical

[1] Aso T. Demography of the menopause and pattern of climacteric symptoms in

conventional hormonal preparations that are considered bioequiv-

the East Asian region. In: First consensus meeting on menopause in the East

alent in practice. More studies are needed to evaluate clinical as

Asian Region. 1997.

well as pharmacokinetic differences between compounded for- [2] Gracia CR, Freeman EW. Acute consequences of the menopausal transition: the

mulations, using various vehicles or administration and dosages. rise of common menopausal symptoms. Endocrinology and Metabolism Clinics

of North America 2004;33:675–89.

The knowledge gained from this study provides the ground-

[3] Iosif CS, Bekassy Z. Prevalence of genito-urinary symptoms in the late

work for future trials to determine the safety and efﬁcacy of

menopause. Acta Obstetricia et Gynecologica Scandinavica 1984;63:257–60.

bioidentical compounded hormones. This successfully conducted [4] Gregersen N, Jensen PT, Giraldi AE. Sexual dysfunction in the peri- and post-

menopause. Status of incidence, pharmacological treatment and possible risks.

randomized, controlled, blinded multi-group study attests to the

A secondary publication. Danish Medical Bulletin 2006;53:349–53.

feasibility of using a similar design in the setting of a larger clinical

[5] Clayton AH, Ninan PT. Depression or menopause? Presentation and manage-

trial. ment of major depressive disorder in perimenopausal and postmenopausal

women. Primary Care Companion to the Journal of Clinical Psychiatry 2010;12. PCC08r00747.

Contributors

[6] Freeman EW. Associations of depression with the transition to menopause.

Menopause 2010;17:823–7.

[7] Woods NF, Mitchell ES. Sleep symptoms during the menopausal transition and

Richa Sood, Lynne Shuster, Darrell Schroeder conceptualized

early postmenopause: observations from the Seattle Midlife Women’s Health

and designed the study. The manuscript was drafted by Richa

Study. Sleep 2010;33:539–49.

Sood, Lynne Shuster, Deborah Rhodes, Dietlind Wahner-Roedler, [8] Landis CA, Moe KE. Sleep and menopause. Nursing Clinics of North America

2004;39:97–115.

Rebecca Bahn. Laboratory services and analysis were carried out

[9] National Institutes of Health. NIH State-of-the-Science Conference Statement

by Ravinder Singh. Pharmaceutical services were provided by Roger

on management of menopause-related symptoms. NIH Consensus and State-

Warndahl. Analysis and interpretation of data were done by Richa of-the Science Statements 2005;22:1–38.

[10] Poniatowski BC, Grimm P, Cohen G. Chemotherapy-induced menopause: a lit-

Sood, Darrell Schroeder, Ravinder Singh, Lynne Shuster, Rebecca

erature review. Cancer Investigation 2001;19:641–8.

Bahn, Deborah Rhodes. Statistical analysis was carried out by Dar-

[11] Hulley S, Grady D, Bush T, et al. Randomized trial of estrogen plus progestin

rell Schroeder. All authors have read and approved the ﬁnal version for secondary prevention of coronary heart disease in postmenopausal women.

of the manuscript. Heart and Estrogen/progestin Replacement Study (HERS) Research Group. Jour-

nal of the American Medical Association 1998;280:605–13.

[12] Rossouw JE, Anderson GL, Prentice RL, et al. Risks and beneﬁts of estrogen

Competing interest plus progestin in healthy postmenopausal women: principal results from the

Women’s Health Initiative randomized controlled trial. Journal of the American

Medical Association 2002;288:321–33.

The authors declare no conﬂict of interest.

[13] Sood R, Shuster L, Smith R, Vincent A, Jatoi A. Counseling postmenopausal

women about bioidentical hormones: ten discussion points for practicing

Funding physicians. Journal of the American Board of Family Medicine 2011;24:202–10.

[14] Jarvinen A, Nykanen S, Paasiniemi L. Absorption and bioavailability of oestradiol

from a gel, a patch and a tablet. Maturitas 1999;32:103–13.

Funding was received by Solvay, an educational grant for

[15] Kuhl H. Pharmacology of estrogens and progestogens: inﬂuence of different

Women’s Health Fellowship, for this article. routes of administration. Climacteric 2005;8(Suppl. 1):3–63.