Reproductive and Developmental Network in Chinese Medicine
Materials and methods
Study Oversight
This
was a multicenter randomized double-blind placebo-controlled trial in
mainland China. Recruited participants were allocated randomly into one
of the three groups in a ratio of 1:1:1. We previously reported details
of the trial protocol (13),
which was designed by the steering committee of the National Clinical
Research Base and approved before initiation by the State Administration
of Traditional Chinese Medicine of the People's Republic of China, the
appointed scientific advisory board. The Institutional Review Boards at
participating hospitals approved the protocol, and every participant
gave written informed consent.
Participants
A
total of 644 women were enrolled in 19 participating sites. Chinese
women with PCOS attempting to get pregnant were eligible if they
fulfilled the following criteria: 1) age 20–40 years; 2) diagnosis of
PCOS according to two of the three Rotterdam 2003 criteria (14),
including oligo-ovulation or anovulation, clinical and/or biochemical
signs of hyperandrogenism, and/or polycystic ovaries; 3) at least one
open fallopian tube and normal uterine cavity documented by
hysterosalpingography, sonohysterography, or diagnostic laparoscopy
within the past 3 years; 4) a male partner with sperm concentration of
15 million/mL and motility of ≥40% in at least one ejaculate; and 5) at
least 1 year of infertility. Subjects were excluded if they used
hormonal drugs or other medications, including Chinese herbal
prescriptions, in the past 3 months; had known severe organ dysfunction
or mental illness; were pregnant, post-miscarriage, postpartum, or
breastfeeding within the past 6 weeks; or had congenital adrenal
hyperplasia, clinically suspected Cushing syndrome, or an
androgen-secreting neoplasm.
Randomization and Blinding
The randomization was performed through a web-based computer program (http://210.76.97.192:8080/cjbyj)
operated by an independent data coordinating center, the Institute of
Basic Clinical Medicine of the China Academy of Chinese Medical
Sciences. The randomization was stratified by the participating sites.
Participants, investigators, physicians taking care of the participants,
laboratory technicians, and data analyzers were blinded to the
assignments.
Study Intervention
After
spontaneous menses or withdrawal bleeding induced by progestin
administration (medroxyprogesterone acetate [Provera; Pfizer Italia],
5 mg/d for 7 days), eligible patients were randomized into one of three
interventions: 1) letrozole and berberine placebo (letrozole group); 2)
berberine and letrozole placebo (berberine group); and 3) letrozole and
berberine (combination group). Each participant received a medication
package on a monthly visit basis that consisted of a monthly supply of
berberine capsules or placebo capsules and one or two packages of pills
(letrozole or letrozole placebo, one package per month for the first
3 months, and two packages per month for the next 3 months). Berberine
or berberine placebo was administrated orally at a daily dose of 1.5 g
for 6 months. Patients received an initial dose of 2.5 mg (one tablet)
of letrozole or one tablet of letrozole placebo on days 3–7 of the first
three treatment cycles. This dose was increased to 5 mg letrozole (two
tablets) or two tablets of letrozole placebo on days 3–7 of the last
three treatment cycles if not pregnant. Induction of withdrawal bleeding
with progestin was scheduled at the discretion of the principal
investigator at each site. All subjects returned monthly on day 22 of
induced or spontaneous menstrual cycles for measurement of serum
progesterone levels and urine hCG test in local laboratories to document
ovulation and pregnancy, received medication packages, and reported
concomitant medication and adverse events. Couples were instructed to
have regular intercourse two to three times a week until achieving
pregnancy. Once participants conceived, they were followed until a
viable intrauterine pregnancy sac was observed (fetal heart motion and
gestational sac visualized on ultrasonography). They were then referred
for antenatal care. Outcomes were tracked through regular interviews
with midwives and abstractions of obstetrical records. Ovulation
detection kits and intrauterine insemination were not used. Berberine
and berberine placebo were produced by Renhetang Pharmaceutical Co.
Letrozole and letrozole placebo were produced by Jiangsu Hengrui
Medicine Co. Neither manufacturer had a financial role in the study.
Fasting
blood samples for assessment of metabolic and hormonal profiles were
drawn at the baseline visit and at the end of the treatment visit at
menstrual cycle days 3–7 and analyzed at the core laboratory in Harbin.
The intra-assay and interassay coefficients of variation of each assay
were <10%. All baseline measures, including assessment of liver and
renal function, were repeated at the end of the visits. Normal values
for the steroids and proteins were 0.6–4.7 nmol/L for P,
0.29–1.67 nmol/L for total T, 46.0–607 pmol/L for E2, 5–10
for free androgen index, 2.4–12.6 mIU/mL for LH, 3.5–12.5 mIU/mL for
FSH, 0.35–4.94 mIU/L for TSH, 102–496 mIU/L for PRL, and 18–114 nmol/L
for SHBG at early follicular stage; and 2.7–24.9 IU/L for fasting
insulin and 0.37–1.47 nmol/L for C-peptide. We did not measure serum
glucose levels because of inaccuracy with long storage of samples in the
core laboratory, resulting in missing homeostasis-model assessment
index data.
Outcome Measures
The
primary outcome was cumulative live births during intervention period.
Secondary outcomes included: 1) ovulation: serum P level >5 ng/mL on
day 22 of each treatment cycle in local sites; 2) conception: a positive
serum or urinary test of hCG; 3) pregnancy: an intrauterine pregnancy
sac with fetal heart motion as determined by means of ultrasonography at
∼8–10 weeks of gestation; 4) multiple pregnancy; 5) pregnancy loss:
loss before 20 completed weeks of intrauterine gestation; 6) other
pregnancy complications, such as stillbirth, gestational diabetes
mellitus, pregnancy-induced hypertension, and small-for-gestational-age
fetus; and 7) adverse events from the study medications. Patients were
asked to record adverse events and to report them to the coordinator at
each visit. Serious adverse events were defined as events that were
fatal or immediately life threatening, that were severely or permanently
disabling, or that required prolonged inpatient hospitalization;
overdoses (intentional or accidental); congenital anomalies; or any
event deemed to be serious by the principal investigator at the study
site.
Statistical Analysis
The
sample size calculation was based on anticipated live birth rate. A
previous meta-analysis suggested that the live birth rate with the use
of letrozole in women with PCOS was ∼22% during a 6-month intervention (13), and our study was designed and completed before a more recently published large randomized trial (6).
We hypothesized that a combination of letrozole and berberine would
increase the live birth rate from 22% to 30%. Accordingly, we estimated
that a sample size of 220 participants per group was required
considering a 20% dropout, 90% power, and an alpha error of 0.05. On the
basis of these assumptions, we needed to enroll 660 subjects for the
study.
Either chi-square test or Fisher exact test was
used at a two-sided significance level of 0.05 for testing differences
among the three study groups for categoric variables. The Kruskal-Wallis
was used to test differences among the three groups for continuous
variables. If significant, a Mann-Whitney U test was used to
test differences between the groups. Kaplan-Meier analysis was used to
compare time to live birth according to treatment groups (5, 6)
body mass index (BMI), hirsutism scores, menstrual patterns, age, and
previous infertility duration. Adverse events were categorized, and the
percentage of patients experiencing adverse events and serious adverse
events in each treatment arm were compared with the use of chi-square
tests. All analyses were performed with the use of SAS software, version
9.2 (SAS Institute). Data were analyzed according to the
intention-to-treat principle.
Results
Study Oversight
It
was approved by the Ethics Committee on April 10, 2009, and registered
in China with identifier ChiCTR-TRC-09000376 on October 8, 2009 (http://apps.who.int/trialsearch/).
The trial was started on October 2009. Owing to the expiration of the
study drug (berberine and matching placebo), the data safety and
monitoring board decided to stop enrollment in November 2013, after 644
patients were enrolled. The first enrollment was on November 11, 2009,
and the last enrollment was February 28, 2013. The first birth occurred
on October 5, 2010, in a patient who became pregnant after she was
enrolled on November 25, 2009. The last live birth occurred on November
12, 2013, in a woman enrolled on December 25th, 2012.
The flowchart of the study is shown in Figure 1.
The number of subjects who withdrew from the study was 16 out of 215
(7.4%) in the letrozole group, 25 out of 214 (11.7%) in the berberine
group, and 15 out of 215 (7.0%) in the combination group (P=.16; Fig. 1). Reasons for withdrawal were similar among the three groups (P=.16 for the three groups; P=.19 for lost to follow-up; P=.88 for drop-out; P=1.0 for protocol violations; and P=.33 for adverse events).
Figure 1
Enrollment flowchart and reproductive outcomes.
There
were no significant differences in the ages of the women, duration of
infertility, BMI, waist/hip circumference, presence of hirsutism,
menstrual pattern, ultrasound features of polycystic ovaries, or
baseline hormonal profile among the three groups (Table 1).
| Biometric feature | Letrozole group (n = 215) | Berberine group (n = 214) | Combination group (n = 215) |
|---|---|---|---|
| Age of women (y) | 27.8 ± 3.6 | 27.8 ± 3.7 | 27.8 ± 3.6 |
| Body mass index (kg/m2) | 24.8 ± 4.5 | 24.5 ± 4.1 | 25.1 ± 5.0 |
| Waist circumference (cm) | 83.5 ± 10.9 | 82.7 ± 11.8 | 83.1 ± 11.8 |
| Hip circumference (cm)a | 98.1 ± 11.4 | 97.7 ± 11.8 | 97.9 ± 10.9 |
| WHR | 0.85 ± 0.08 | 0.85 ± 0.08 | 0.85 ± 0.07 |
| Hirsutism (Ferriman-Gallwey ≥5) | 78/202 (38.6) | 76/201 (37.8) | 60/209 (28.7) |
| Menstrual patterna | |||
| Oligomenorrhea | 126/201 (62.7) | 116/193 (60.1) | 121/198 (61.1) |
| Regular menses | 75/201 (37.3) | 77/193 (39.9) | 77/198 (38.9) |
| Duration of infertility (mo) | 32.7 ± 24.0 | 28.5 ± 21.6 | 29.8 ± 21.3 |
| Previous infertility therapy | 178 (82.8) | 180 (84.1) | 185 (86.0) |
| Traditional Chinese medicine | 107/178 (60.1) | 112/180 (62.2) | 110/185 (59.5) |
| Ovulation drugs | 121/178 (68.0) | 109/180 (60.6) | 107/185 (57.8) |
| Assisted reproductive technology | 9/178 (5.1) | 11/180 (6.1) | 12/185 (6.5) |
| Other therapies | 8/178 (4.5) | 8/180 (4.4) | 6/185 (3.2) |
| Previous pregnancyb | |||
| Conception | 67 (31.2) | 62 (29.0) | 80 (37.2) |
| Live birth | 9 (4.2) | 7 (3.3) | 7 (3.3) |
| Miscarriage | 13 (6.1) | 16 (7.5) | 26 (12.1) |
| Termination of pregnancy | 48 (22.3) | 43 (20.1) | 64 (29.8) |
| Ultrasonographic findings | |||
| Polycystic ovary morphology | 104/151 (68.9) | 113/153 (73.9) | 98/155 (63.2) |
| Ovarian volume (cm3) | |||
| Left ovary | 10.3 ± 6.8 | 9.5 ± 7.9 | 11.0 ± 6.8 |
| Right ovary | 11.0 ± 6.3 | 10.2 ± 6.4 | 11.6 ± 7.4 |
| Fasting serum levels | |||
| TSH (mIU/L) | 2.6 ± 1.9 | 2.6 ± 1.5 | 2.7 ± 1.8 |
| PRL (mIU/L) | 318.0 ± 160.9 | 311.3 ± 168.4 | 296.7 ± 155.8 |
| LH (mIU/L) | 10.9 ± 6.7 | 10.0 ± 6.5 | 10.4 ± 6.1 |
| FSH (mIU/L) | 5.7 ± 2.1 | 5.3 ± 1.8 | 5.5 ± 1.9 |
| LH/FSH | 2.0 ± 1.1 | 1.9 ± 1.2 | 1.9 ± 1.1 |
| P (nmol/L) | 6.2 ± 13.1 | 5.7 ± 14.9 | 6.6 ± 15.6 |
| E2 (pmol/L) | 213.9 ± 206.3 | 244.0 ± 252.4 | 228.8 ± 233.0 |
| T (nmol/L) | 1.6 ± 0.8 | 1.4 ± 0.7 | 1.4 ± 0.7 |
| SHBG (nmol/L) | 48.1 ± 31.1 | 47.6 ± 32.9 | 42.6 ± 28.3 |
| Free androgen indexb | 5.0 ± 4.8 | 5.3 ± 6.3 | 5.1 ± 5.8 |
| Insulin (mIU/L) | 11.9 ± 7.7 | 11.1 ± 8.3 | 12.3 ± 8.5 |
| Peptide C (nmol/L) | 0.62 ± 0.45 | 0.55 ± 0.44 | 0.59 ± 0.35 |
Note: Data are presents as mean ± SD or n (%).
aMissing information in some subjects.
bThe free androgen index was calculated according to the following formula: (total T [nmol/L]/SHBG [nmol/L]) × 100.
Primary Outcomes
A
total of 199 live births occurred, including 195 singletons and four
sets of twins. The rate of cumulative live births was similar between
the letrozole and combination groups (36.3% vs. 34.4%; odds ratio [OR]
0.95, 95% confidence interval [CI] 0.73–1.23; P=.687), and significantly higher in those groups together than in the berberine group (22%; OR 1.68, 95% CI 1.23–2.28 [P=.001]; OR 1.57, 95% CI 1.15–2.14 [P=.004]; respectively; Table 2; Supplemental Fig. 1A [available online at www.fertstert.org).
Birth weight among live births was similar among the three groups.
There were three twin live births in the combination group, one in the
letrozole group, and none in the berberine group.
| Outcome | Letrozole group (n = 215) | Berberine group (n = 214) | Combination group (n = 215) | OR (95% CI) between combination and letrozole | P value | OR (95% CI) between combination and berberine | P value | OR (95% CI) between letrozole and berberine | P value |
|---|---|---|---|---|---|---|---|---|---|
| Primary outcomes | |||||||||
| Live birth | 78/215 (36.3) | 47/214 (22.0) | 74/215 (34.4) | 0.95 (0.73–1.23) | .687 | 1.57 (1.15–2.14) | .004 | 1.68 (1.23–2.28) | .001 |
| Singleton live birth | 77/78 (98.7) | 47/47 (100) | 71/74 (95.9) | 0.97 (0.92–1.03) | .357 | 0.96 (0.92–1.01) | .28 | 0.99 (0.96–1.01) | 1.000 |
| Twin live birth | 1/78 (0.1) | 0/47 (0) | 3/74 (0.4) | 3.16 (0.34–29.72) | .357 | 4.48 (0.24–84.82) | .28 | 1.82 (0.08–43.85) | 1.000 |
| Birth weight (g), mean ± SD | 3,463 ± 575 | 3,542 ± 399 | 3,484 ± 504 | 21.20 (−177.16–219.57) | .845 | −58.23 (−248.84–132.38) | .246 | −79.43 (−282.07–123.20) | .216 |
| Secondary outcomes | |||||||||
| Ovulation | 473/796 (59.4) (59.4) | 302/831 (36.3) | 486/797 (61.0) | 1.03 (0.95–1.11) | .526 | 1.68 (1.51–1.87) | <.0001 | 1.64 (1.47–1.82) | <.0001 |
| Conception | 98/215 (45.6) | 61/214 (28.5) | 105/215 (48.8) | 1.07 (0.88–1.31) | .499 | 1.71 (1.33–2.21) | <.0001 | 1.60 (1.24–2.07) | .0003 |
| Immediate loss to follow-up | 3/98 (3.1) | 0 | 4/105 (3.8) | 1.24 (0.29–5.42) | 1.000 | – | – | – | – |
| Pregnancy | 84/215 (39.1) | 48/214 (22.4) | 81/215 (37.7) | 0.96 (0.76–1.23) | .766 | 1.68 (1.24–2.27) | .0006 | 1.741 (1.29–2.35) | .0002 |
| Singleton | 83/84 (98.8) | 48/48 (100.0) | 78/81 (96.3) | 0.97 (0.93–1.02) | .361 | 0.96 (0.92–1.01) | .294 | 0.99 (0.97–1.01) | 1.000 |
| Twins | 1/84 (1.2) | 0 | 3/81 (3.7) | 3.11 (0.33–29.30) | .361 | – | – | – | – |
| Pregnancy loss | 17/98 (17.4) | 14/61 (23.0) | 27/105 (25.7) | 1.48 (0.86–2.55) | .148 | 1.12 (0.64–1.97) | .691 | 0.76 (0.40–1.42) | .386 |
| In the 1st trimester | 13/98 (13.2) | 14/61 (23.0) | 24/105 (22.9) | 1.72 (0.93–3.19) | .077 | 0.996 (0.56–1.78) | .989 | 0.58 (0.29–1.15) | .114 |
| In the 2nd trimester | 4/98 (4.1) | 0 | 3/105 (2.9) | 0.70 (0.16–3.05) | .714 | – | – | – | – |
| Fecundity among ovulated cycles | |||||||||
| Conception | 98/473 (20.7) | 61/302 (20.2) | 105/486 (21.6) | 1.05 (0.77–1.44) | .737 | 1.09 (0.76–1.55) | .638 | 1.03 (0.72–1.48) | .861 |
| Pregnancy | 84/473 (17.5) | 48/302 (15.9) | 81/486 (16.1) | 0.94 (0.71–1.24) | .654 | 1.05 (0.76–1.45) | .776 | 1.13 (0.76–1.66) | .549 |
| Live birth | 78/473 (16.3) | 47/302 (15.6) | 74/486 (14.6) | 0.92 (0.69–1.24) | .592 | 0.98 (0.70–1.37) | .899 | 1.06 (0.71–1.57) | .791 |
| Fecundity among subjects who ovulated | |||||||||
| Conception | 98/188 (52.1) | 61/147 (41.5) | 105/184 (57.1) | 1.09 (0.91–1.32) | .339 | 1.38 (1.09–1.73) | .005 | 1.26 (0.99–1.59) | .053 |
| Pregnancy | 84/188 (44.2) | 48/147 (32.7) | 81/184 (42.4) | 0.99 (0.78–1.24) | .898 | 1.35 (1.02–1.79) | .035 | 1.37 (1.03–1.81) | .025 |
| Live birth | 78/188 (41.0) | 47/147 (32.0) | 74/184 (38.6) | 0.97 (0.76–1.24) | .803 | 1.26 (0.94–1.69) | .122 | 1.30 (0.97–1.74) | .074 |
Note:
Ovulation was defined as a serum P level according to the standard of
the local laboratory (minimum value of luteal phase) or >5 ng/mL.
Conception was defined as any positive serum level of hCG. Pregnancy was
defined as an intrauterine pregnancy sac with fetal heart motion as
determined by ultrasonography. Live birth was defined as the delivery of
a viable infant.
Independently from
treatment, subjects with age <33 years had significantly higher rates
of live births than did women whose age was >33 years (P=.018). There were no significant differences in live birth rate stratified by BMI (P=.782), menstrual cycle pattern (P=.689), hirsutism ,or duration of infertility (data not shown; Supplemental Fig. 1B–1D).
Secondary Outcomes
The
rates of ovulation, conception, and pregnancy also were similar between
the combination and letrozole groups, and significantly higher in those
groups than in the berberine group. Regarding fecundity among subjects
who ovulated, the letrozole and combination groups were superior to the
berberine group in terms of conception rates and pregnancy rates.
However, there were no differences among the three groups in rates of
conception, pregnancies, or live births among ovulatory cycles during
treatment. The three groups had similar rates of pregnancy loss after
conception. The within-group changes in BMI and waist circumference
between last visit and baseline were significant or marginal in the
berberine group (−0.31 ± 4.34 kg/m2 [P=.0017]; −0.45 ± 4.31 cm [P=.0553]) and not significant in the letrozole group (0.65 ± 10.73 kg/m2 [P=.0942]; 0.82 ± 13.13 cm [P=.0715]) or the combination group (−0.08 ± 2.23 kg/m2 [P=.241]; −0.08 ± 4.97 cm [P=.2703]), although the between-group comparisons were not significant among three groups (P=.45).
Adverse Events and Pregnancy Complication
No
serious adverse events occurred during the intervention period in any
of the three groups. Berberine was associated with a significantly
higher incidence of constipation and nausea, and letrozole was
associated with a significantly higher incidence of fatigue and hot
flashes. There were no significant differences among the three groups
regarding rates of total cases of adverse events (37.0%, 48.0%, and
45.5% for the letrozole, berberine, and combination groups,
respectively) or cases of serious adverse events, including ectopic
pregnancies, pregnancy loss during the second trimester, or preterm
labor (21.4%, 20.8%, and 23.5%, respectively; Table 3).
| Adverse event | Letrozole group (n = 215) | Berberine group (n = 214) | Combination group (n = 215) |
|---|---|---|---|
| Serious adverse event from study medication | 0/215 (0) | 0/214 (0) | 0/215 (0) |
| Other adverse events | |||
| Constipation | 10/215 (4.7) | 26/214 (12.1)a | 12/215 (5.6) |
| Nausea | 13/215 (6.0) | 34/214 (15.9)a | 26/215 (12.1) |
| Diarrhea | 13/215 (6.0) | 5/214 (2.3) | 5/215 (2.3) |
| Hot flashes | 21/215 (9.8) | 6/214 (2.8)a | 19/215 (8.8) |
| Fatigue | 17/215 (7.9) | 5/214 (2.3)a | 18/215 (8.4) |
| Serious events during pregnancy | |||
| First trimester | |||
| Ectopic pregnancy | 2/84 (2.4) | 2/48 (4.2) | 2/81 (2.5) |
| Second and third trimesters | |||
| Pregnancy loss after 12 wk | 4/84 (4.8) | 0/48 (0) | 3/81 (3.7) |
| Preeclampsia | 4/84 (4.8) | 6/48 (12.5) | 7/81 (8.6) |
| Gestational diabetes | 3/84 (3.6) | 3/48 (6.3) | 7/81 (8.6) |
| Preterm labor | 2/84 (2.4) | 0/48 (0) | 2/81 (2.5) |
| Premature rupture of membranes | 5/84 (6.0) | 1/48 (2.1) | 2/81 (2.5) |
| Serious events in fetus and infant | |||
| Fetal abnormality | 1/84 (1.2) | 0/48 (0) | 0/48 (0) |
| Neonatal death | 0/78 (0) | 0/47 (0) | 1/74 (1.4) |
aP<.05 versus other two groups.
During
pregnancy, the most common complication was pregnancy-induced
hypertension, followed by gestational diabetes, threatened abortion, and
premature rupture of membranes. There were no significant differences
among the three treatment groups for these events. Two major congenital
anomalies were reported. One fetal abnormality in the letrozole group
was detected as hydrocephalus by ultrasound examination at gestational
week 16, resulting in termination of the pregnancy without further
autopsy. Another abnormality in the combination group came to an infant
death on day 31 after birth by parent's refusal of further treatment for
a major ventricular septal defect and pulmonary stenosis.
Discussion
Our
findings did not support the hypothesis that a combination of letrozole
and berberine is superior to letrozole alone for achieving live birth
in infertile women with PCOS. The rates of ovulation, conception,
pregnancy, and live birth were similar between the letrozole and
combination groups. These end points were significantly higher in both
groups with letrozole than in the berberine alone group. We found that
the higher ovulation rate per cycle accounted for the superiority of
this cumulative live birth rate for letrozole or combination treatment
compared with berberine alone.
The use of
insulin-sensitizing agents such as metformin in women with PCOS
undergoing ovulation induction has been widely studied. A Cochrane
review (15)
showed that there was no evidence that metformin improved live birth
rates, whether used alone or in combination with clomiphene. Therefore,
the role of metformin in improving reproductive outcomes in women with
PCOS appears to be limited. Berberine, an active ingredient from Chinese
medicinal herbs, has multiple biologic activities and pharmacologic
effects in several metabolic diseases, such as type 2 diabetes mellitus,
hyperlipidemia, and nonalcoholic fatty liver disease (7, 8, 16).
A systemic review and meta-analysis for berberine in the treatment of
type 2 diabetes mellitus demonstrated glycemic control similar to other
oral hypoglycemic agents. However, berberine also additionally showed an
antidyslipidemic effect (7).
The relevant targets of berberine might link to the insulin pathway,
adenosine monophosphate–activated protein kinase signaling, the gut
environment, and hepatic lipid transportation (16).
In a study comparing berberine and metformin (12),
89 Chinese women with PCOS and insulin resistance were randomized into
one of three treatment groups: berberine and cyproterone acetate,
metformin and cyproterone acetate, or placebo and cyproterone acetate
for three months. Berberine showed similar restoration of insulin
sensitivity and reduction of hyperandrogenemia compared with metformin.
Berberine also appeared to have a greater effect on changes in body
composition and dyslipidemia in PCOS patients.
Our
results showed that ovulation and live birth rates for a combination of
letrozole and berberine were similar to letrozole alone. However, there
were no differences among the three groups in rates of conception,
pregnancies, or live births when only ovulatory cycles were considered.
This implied that improvement in the metabolic profile by berberine did
not affect the ovulation or live birth rates achieved with letrozole. We
did not measure insulin sensitivity before and after exposure to the
study medication in the present study. We did not find significant
differences in live birth rates stratified by BMI, hirsutism score,
menstrual patterns of regularity or oligomenorrhea, and duration of
infertility. Our results contrast with those from a study conducted in
the Netherlands (17)
where BMI, age, free androgen index, and cycle history were all
associated with live birth in women with PCOS receiving clomiphene. In
the population-based Northern Finland Birth Cohort 1966 study (18),
previous oligo-amenorrhea and/or hirsutism and obesity were both found
to be independently associated with decreased fecundity. These
differences might be related to the distinct phenotypes of PCOS, in that
a Chinese cohort is more likely to have less hyperandrogenism and to be
more lean than a European cohort (1, 2).
Letrozole
has a shorter half-life than clomiphene, leading to a shorter exposure
during implantation and early fetal development and, therefore, a low
theoretical risk of congenital anomalies in the offspring (6, 19).
There has been significant concern for congenital anomalies since the
introduction of letrozole for use in inducing ovulation. In the present
trial, we reported one fetal abnormality in the letrozole group, one
neonatal death in the combination group, and no abnormalities in the
berberine group. However, our study was underpowered to detect a
significant difference for rare but potentially serious adverse events.
In the PCOS II trial (6),
Legro et al. found four major congenital anomalies in the letrozole
group and one in the clomiphene group. There was no pattern to the four
major anomalies with letrozole, implying that they were random events.
The rate of congenital anomalies in the present study is below what has
been reported in other studies (20).
In
the present study, the use of berberine alone achieved a 36% ovulation
rate per cycle, similar to metformin, and a 22% cumulative live birth
rate, similar to clomiphene, after 6 months of use (5).
To the best of our knowledge, this is the first study to show the
effectiveness of berberine alone on the ovulation and live birth rates
in women with PCOS. However, we can not make conclusions regarding
efficacy because berberine was not compared with a placebo or no
treatment. A study directly comparing metformin and berberine for
ovulation induction is worth performing in the future.
One
of the limitations of the present study was its early termination
because of the expiration of the study drug (berberine and its placebo).
However, the absolute difference (12%–14%) in live birth rates for the
original two primary comparisons of letrozole and berberine in this
trial was greater than the projected 8%; therefore, the unexpected
reduction to 644 participants still provided adequate power (≥90%) for
the study. Other limitations included not all subjects having pelvic
sonograms to assess for polycystic ovary morphology and the inclusion of
subjects with regular cycles for ovulation. The major strength of this
study is that it was a large multicenter double-blind trial with close
monitoring of adverse and serious adverse events and tracking of live
births in line with our recent Harbin Consensus (21). Also, it demonstrates the high live birth rate achieved with letrozole in a population with a relatively lower BMI range.
In
summary, berberine did not add fecundity in infertility conditions with
PCOS when used in combination with the new ovulation agent letrozole.
Participant list
National
Reproductive and Developmental Network in Chinese Medicine for this
trial included Li-Hui Hou, M.D., Department of Obstetrics and
Gynecology, First Affiliated Hospital, Heilongjiang University of
Chinese Medicine, Harbin, China; Yong-Yan Wang, M.D., Feng Tian, Ph.D.,
and Yan-Ming Xie, M.D., Institute of Basic Clinical Medicine, China
Academy of Chinese Medical Science, Beijing, China; Jin-Feng Zhang,
M.D., Department of Obstetrics and Gynecology, Shanxi Province Hospital
of Chinese Medicine, Taiyuan, China; Ya-Qin Gao, M.D., Center for
Reproductive Medicine, Daqing Oilfield General Hospital, Daqing, China;
Shao-Min Du, M.D., Department of Obstetrics and Gynecology, Daqing
Longnan Hospital, Daqing, China; Ying Yan, M.D., Department of
Gynecology, First Affiliated Hospital, Tianjin University of Chinese
Medicine, Tianjin, China; Pei-Lin Li, M.D., Department of Gynecology,
Second Affiliated Hospital, Tianjin University of Chinese Medicine,
Tianjin, China; Jin-Ying Fu, M.D., Department of Obstetrics and
Gynecology, Second Affiliated Hospital, Henan College of Chinese
Medicine, Zhengzhou, China; Wei-Li Li, M.D., Department of Obstetrics
and Gynecology, Affiliated Hospital, Anhui University of Chinese
Medicine, Hefei, China; Zhen-Yu Tan, M.D., Department of Obstetrics and
Gynecology, First Affiliated Hospital, Hunan University of Chinese
Medicine, Changsha, China; Feng-Jie He, M.D., Department of Obstetrics
and Gynecology, Affiliated Hospital, Shanxi College of Chinese Medicine,
Xianyang, China; Cai-Fei Ding, M.D., Center for Reproductive Medicine,
Zhejiang Province Hospital of Integrative Medicine, Hangzhou, China;
Xiao-Bin Li, M.D., Department of Obstetrics and Gynecology, Guangdong
Province Hospital of Chinese Medicine, Guangzhou, China; Xian-Ji Shen,
M.D., Outpatient Department, Mudanjiang Maternal and Children's Health
Hospital, Mudanjiang, China; Mu-Er An, M.D., Department of Obstetrics
and Gynecology, First Affiliated Hospital, Harbin Medical University,
Harbin, China; Guang-Zhu Yu, M.D., Department of Gynecology, Suqian
Maternal and Children's Health Hospital, Suqian, China; Robert M.
Silver, M.D., Department of Obstetrics and Gynecology, University of
Utah Health Sciences Center, Salt Lake City, Utah; and Elisabet
Stener-Victorin, Ph.D., Department of Physiology and Pharmacology,
Karolinska Institutet, Stockholm, Sweden.
The members of
the Steering Committee for this network were Xiao-Ke Wu, Jian-Ping Liu,
Elisabet Stener-Victorin, Ernest H. Y. Ng, Zhaoling You, and Chengzong
Xiao. The members of the Scientific Advisory Board were Drs. Sulun Sun,
Weiliang Weng, Yanming Xie, Yixun Liu, Yongyan Wang for the National
Clinical Research Base of Chinese Medicine in this network at the First
Affiliated Hospital, Heilongjiang University of Chinese Medicine,
Harbin, People's Republic of China.
Acknowledgment
The
authors thank all of the site participants and the staff in the Harbin
Administrative office for their contributions, including Drs. Hongying
Kuang, Yan Li, Hongli Ma, Wenjuan Shen, and Jingshu Gao during their
domestic and international training for the trial.
Supplementary data
