Methotrexate (50 mg/m2 intramuscularly and 50 mg
orally) followed by vaginal misoprostol have proven to
be > 90% effective at causing abortion in women at less
than 49 days’gestation. Although the effectiveness of the
oral dose (which has a lower serum bioavaiIability)
demonstrates that a methotrexate dose of 50 mglm’may
be more than necessary, an intramuscular regimen is
more advantageous because it is less costly. This trial
was designed to investigate the potential effectiveness of
a single dose of methotrexate, 75 mg intramuscularly, in
a regimen for early abortion. One hundred subjects
received 75 mg methotrexate intramuscularly followed 5
to 6 days later by 800 pg misoprostol vaginally. The
misoprostol dose was repeated if the abortion did not
occur. Outcome measures included successful abortion
(complete abortion without requiring a surgical proce-
dure), duration of vaginal bleeding, and side effects. One
subject was lost to Jollow-up. Complete abortion oc-
curred in 94 of 99 (94.9%, 95% CI 90.6, 99.3%) patients.
The complete abortion rate was no different for earlier
gestations: 38 of 40 (95.0%, 95% CI 88.2, 100%) at up to
42 days’ gestation and 56 of 59 (94.9%, 95% CI 89.3,
100%) at more than 42 days’ gestation (p = 0.99).
Abortion occurred in the 24 h following the initial or
repeat misoprostol dose (immediate success) in 70.7%;
the remaining 24.2% of women who aborted did so after
a delay of 22 t 10 days (mean + standard deviation).
Vaginal bleeding lasted 17 -C 8 days and I1 ? 7 days in
immediate success and delayed success patients, respec-
tively. Overall, 77.8%, 86.9%, and 91.9% of patients had
passed the pregnancy by 14, 28, and 35 days, respec-
tively, after receiving methotrexate. This preliminary
evaluation demonstrates that a medical abortion regi-
men using 75 mg methotrexate intramuscularly appears
to have similar effectiveness to one with 50 mg/m2
Introduction
I ntramuscular methotrexate (50 mg/m2) followed
by vaginal misoprostol has been shown to be
> 90% effective in causing an abortion in pregnan-
cies up to 49 days’ gestation.’ Similar effectiveness
has recently been shown by Creinin and coworkers
when the methotrexate was administered orally in a
single dose of 50 mg to all patients regardless of body
surface area.2 Because the bioavailability of oral
methotrexate is approximately 15% less than that of
an equivalent dose administered intramuscularly,3,4
the dose of 50 mg orally is equivalent to 42.5 mg
intramuscularly.
In the study by Creinin and coworkers2 the average
body surface area was 1.74 m2 translating to an
average oral dose of methotrexate administered of
28.7 mg/m2 (range, 21.2 to 36.2 mg/m2). This dose
would be equivalent to an intramuscular dose of 24.4
mg/m2 (range, 18.0 to 30.8 mg/m2), suggesting that a
dose of 50 mg/m2 intramuscularly is more than nec-
essary.
The main advantage of intramuscular over oral
methotrexate is cost. At my institution, a 50 mg oral
dose costs approximately $20 as compared with $4 for
the appropriate dose of parenteral methotrexate. This
trial was designed to study a single dose of intramus-
cular methotrexate (75 mg) to be given to all patients.
A regimen using the same dose for all patients would
simplify the treatment by eliminating the need to
calculate the body surface area and adjust the dose on
an individual basis. The dose of 75 mg was chosen
because: 1) methotrexate is available in vials of 25
mg/mL (as well as larger multidose vials) and a dose of
75 mg is an even 3 mL; 2) the average body surface
area in prior trials 1,5-7 was approximately 1.71 to 1.72
m2, which translates to a dose of approximately 8
mg; because this may be a higher dose than necessary,
a slightly lower dose is justified for the study; 3) very
few women were less than 1 SO m2 in the prior trials,
which translates to a dose of 7.5 rng;lJ5-’ therefore,
few, if any, women would receive a dose greater than
what they would receive if the dosing was based on
body surface area.
Methods
Healthy, English-speaking women were recruited
for a prospective trial approved by the Institutional
Review Board of Magee-Womens Hospital. Entry
criteria included: 1) 18 years of age or older, 2)
requesting an elective abortion, 3) intrauterine
pregnancy not exceeding 49 days’ gestation docu-
mented by vaginal ultrasound, 4) willing to abstain
from intercourse and alcohol for the first 14 days of
the study and comply with the visit schedule, 5)
adequate venous access for multiple phlebotomies,
and 6) access to a telephone. Exclusion criteria
included: 1) use’ of prenatal vitamins or any other
medications containing folate, 2) hemoglobin < 10
g/dL, 3) aspartate aminotransferase more than twice
normal or active liver disease, 4) serum creati-
nine > 1.5 mg/dL or active renal disease, 5) active
inflammatory bowel disease, 6) uncontrolled sei-
zure disorder (> 1 seizure per week), and 7) known
intolerance of or allergy to methotrexate or miso-
prostol.
Patients who enrolled signed an informed consent
and agreed to suction abortion should the pregnancy
be viable (defined as the presence of cardiac activity
on vaginal ultrasound) 14 days after initiating the
study. Vaginal ultrasound was performed and gesta-
tional age (GA) estimated by mean sac diameter or, if
an embryonic pole was present, by crown-rump
length using described criteria.’ Estimated gesta-
tional age was based on last menstrual period (LMP)
but was changed if the ultrasound estimate differed
by 4 days or more from the gestational age by LMP.*f’
On the day of the consent, a history and physical
examination, baseline hemoglobin, and blood type
were performed. Serum aspartate aminotransferase
and creatinine were obtained if clinically indicated
for evaluation of potential liver or renal dysfunction.
If all of the laboratory requirements for study entry
were met, 75 mg methotrexate was injected intra-
muscularly at least 24 h after obtaining consent as is
required by the Pennsylvania Abortion Control Act;
this was considered study day 1. If the patient’s blood
type was Rh-negative, she also received 50 kg Rh-
immune globulin intramuscularly. Before methotrex-
ate administration, a quantitative serum B-hCG (B-
human chorionic gonadotropin) was collected.
Subjects were given four tablets of 200 pg misopr-
ostol to take home and instructions to place the
tablets as high up as possible in the vagina on study
day 6 or 7. All patients were given a prescription for
20 tablets of acetaminophen with codeine phosphate
(300 mg/30 mg) or acetaminophen with hydrocodone
(500 mg/5 mg). Patients were instructed to use ibu-
profen or acetaminophen initially and to use the
prescribed narcotic only if necessary. Patients were
also given a list of foods lo high in folate to avoid
eating for the first 2 weeks of the study or until the
abortion had occurred, whichever came first. Patients
were not required or instructed to lie down after
misoprostol administration.
The subjects were also instructed to notify the
researchers if heavy vaginal bleeding occurred before
administering the misoprostol tablets. These patients
came into the office for a vaginal ultrasound; if the
ultrasound demonstrated a gestational sac, the sub-
ject was told to return home and administer the
misoprostol as directed. If the ultrasound showed no
gestational sac, the subject was instructed to return
on day 36 for follow-up.
Subjects otherwise returned on day 8, at which
time the subject was questioned as to side effects that
occurred after the methotrexate and misoprostol and
the date and time of misoprostol administration. A
vaginal ultrasound was performed; if the gestational
sac was absent, the subject was instructed to return in
approximately 4 weeks for follow-up. If the gesta-
tional sac was still present, 800 kg misoprostol was
administered by the clinician. The technique of mi-
soprostol administration, using four 200 p,g tablets,
has been described previously.5,7 If embryonic cardiac
activity was not present, the subject was instructed to
return on day 36. If embryonic cardiac activity was
present, the subject returned on day 15.
For subjects who were asked to return on day 15, a
history of events since day 8 was obtained and a
vaginal ultrasound performed. If the gestational sac
and embryonic cardiac activity were present, a surgi-
cal abortion was performed. If the gestational sac was
absent or still present and without cardiac activity,
the patient returned in 3 weeks (day 36).
When subjects returned on day 36, a history of
events since the last visit was obtained. A vaginal
ultrasound was performed if the ultrasound examina-
tion at the last visit had not demonstrated passage of
the gestational sac. If the ultrasound on day 36
showed the gestational sac still to be present, the
patient was offered intervention (another dose of
misoprostol or a surgical abortion); if the woman
declined intervention at this time, then she was
followed at 1 to 2 week intervals until she passed the
pregnancy or requested a surgical abortion. All sub
ects were permitted at any time to request a surgical
procedure rather than continuing to wait for passage.
Subjects not started on hormonal contraception were
instructed to contact the investigators when menses
returned; alternatively, the researchers contacted the
study subjects at regular intervals to obtain this
information.
A questionnaire was administered to each patient
before methotrexate administration and after the
study was completed; these questions investigated
each woman’s reasons for wanting an abortion, why
she wanted a medical abortion, her prior experience
with surgical abortion, and her experience during the
study with medical abortion.
Before the study began, the following outcomes
were defined: 1) any patient who passed the preg-
nancy after methotrexate administration and before
day 4 was considered to have a spontaneous abortion
not as a result of the methotrexate; new subjects were
recruited to replace these patients; 2) an abortion
would be considered successful if complete abortion
occurred without requiring a surgical procedure; 3)
immediate success would be a complete abortion in
which the pregnancy passed after day 3 without
misoprostol administration or during the 24 h after
the initial or repeat dose of misoprostol; 4) delayed
success would be a complete abortion that occurred
more than 24 h after the repeat dose of misoprostol;
and 5) any patient who had not passed the pregnancy
by day 36 and received any additional misoprostol
was considered a treatment failure. The day of the
abortion for the delayed success group was considered
to be the day heavy bleeding started. The diagnosis of
incomplete abortion was to be based on persistent
heavy vaginal bleeding and the clinical impression of
the principal investigator.
Serum P-hCG was measured by chemilumines-
cence (Ciba-Corning, Medfield, MA) using the Third
International Reference Preparation. Body surface
area was calculated using a body surface area table.’ ’
Because this was a preliminary, noncomparative
analysis of a new dosing regimen for medical abor-
tion, a sample size calculation was not appropriate.
Prior studies report treatment success rates of approx-
imately 2 90% up to 49 days’ gestation. If similar
success rates apply with the new dosing regimen, a
sample of 100 subjects would allow an estimate of the
true success rate within a margin of sampling error
of 2 5.9 percentage points.
Statistical analyses were performed using the SAS
statistical package (SAS Institute, Cary, NC). Statis-
tical analyses of treatment success and side effects
was performed using Fisher’s exact and Wilcoxon
nonparametric rank tests when appropriate. A p = 0.5
Patient characteristics are presented in Table 1. A
subject who received her methotrexate at 42 days’
gestation was lost to follow-up after her day 8 visit.
The ultrasound showed no embryonic cardiac activity
at that time. Ninety-seven (96.9%) subjects had a
body surface area (BSA) > 1.50 m2; these subjects,
therefore, received a lower intramuscular dose of
methotrexate then they would have based on a dose of
50 mg/m2.
Abortion occurred without need for a surgical pro-
cedure in 94 of 99 (94.9%, 95% CI 90.6, 99.3%).
Immediate success abortion occurred in 70 (70.7%,
95% CI 61.7, 79.7%) women. A total of 27 and 43 of
these women were 35 to 42 days’ gestation and 43 to
49 days’ gestation, respectively. All but two of the
immediate success patients in each group aborted
after the first dose of misoprostol. The 24 women who
aborted without any further intervention did so after
a delay of 22 ? 10 days (median 19 days, range 10 to
42 days). The complete abortion rate did not differ by
gestation: 38 of 40 (95.0%, 95% CI 88.2, 100%) at up
to 42 days’ gestation and 56 of 59 (94.9%, 95% CI
89.3, 100%) at more than 42 days’ gestation (p = 0.99).
because of the presence of embryonic cardiac activity
on day 8.
Five subjects required surgical intervention. Two
ultrasound examinations on study day 15 still dem-
onstrated embryonic cardiac activity (subjects ini-
tially 46 and 49 days’ gestation with body surface
areas of 1.80 and 1.90 m2, respectively). One subject
(41 days’ gestation) had not passed the pregnancy by
day 36. Two subjects (42 and 46 days’ gestation) had
incomplete abortions. One of these women presented
to the Emergency Department on the day after receiv-
ing a repeat dose of misoprostol asking for her private
physician. She wanted to have a surgical abortion and
did not tell her physician that she was participating in
the study. No subjects required surgical intervention
for hemorrhage. Table 2 summarizes the abortion rate
by study day.
Information on cramping after the first dose of
misoprostol was available for 98 subjects; eight sub-
jects experienced none and one had missing data.
Cramping in the remaining 89 women began within
3.2 + 2.4 h (median 3.0 h). Bleeding began within
5.9 + 4.0 h (median 5.0 h) in 92 women after the
initial dose of misoprostol; five women experienced
no bleeding and two had missing data. The time for
onset of cramping in patients who successfully
aborted after the first dose of misoprostol(3.4 + 2.6 h,
median 3.5) was not different from that of women
who did not (2.8 + 2.0 h, median 3.0, p = 0.31).
Similarly, the time for onset of bleeding in patients
who successfully aborted after the first dose of miso-
prostol (5.2 2 2.8 h, median 4.6) was not different
than that of women who did not (7.6 ? 5.5 h, median
6.5) although it almost reached statistical significance
(p = 0.09).
Information on the duration of vaginal bleeding and
spotting was available for 69 of 70 (98.6%) immeccess and 23 of 24 (95.8%) delayed success patients.
Total vaginal bleeding in immediate and delayed
success patients lasted 17 ? 8 days (median 15 days,
range 4 to 37 days) and 11 + 7 days (median 11 days,
range 3 to 37 days), respectively. No patient required
a transfusion.
Following methotrexate, nausea occurred in 47%,
vomiting in 12%, diarrhea in 22%, headache in 16%,
dizziness in 21%, and warmth/hot flashes in 43%;
22% also stated that they experienced some cramping
between the methotrexate and misoprostol. No sub-
ject complained of oral ulcers. After misoprostol,
nausea occurred in 27%, vomiting in 13%, diarrhea in
26%, dizziness in 18%, headache in 14%, and subjec-
tive fever or chills in 49% of subjects.
Pain from uterine cramping did not require medi-
cation in 18 of 99 (18.2%) patients; all of these women
successfully aborted. All of the remaining 8 1 patients
used only oral medication; three (3.7%) of these
patients stated they did not have adequate relief of
pain with these medications.
A nonhormonal method of contraception was cho-
sen by 49 (49.5%) patients. Five of these women were
lost to follow-up; menses occurred in the remaining
women 37 -+ 8 days (median 36 days, range 24 to 63
days) after the abortion. Because not all patients had
regular menstrual cycles before becoming pregnant,
this wide range for the time to return of menses is
expected.
Discussion
This initial investigation of a single dose of metho-
trexate for all patients having a medical abortion up
to 49 days’ gestation demonstrates that this simpler
treatment regimen appears to be as effective as a
calculated intramuscular dose (50 mg/m2)l and an
oral dose of 50 mg.2 The dose of 75 mg is less than This initial investigation of a single dose of metho-
trexate for all patients having a medical abortion up
to 49 days’ gestation demonstrates that this simpler
treatment regimen appears to be as effective as a
calculated intramuscular dose (50 mg/m2)l and an
oral dose of 50 mg.2 The dose of 75 mg is less than uld have been administered based on surface area
for all but three of the 99 subjects. The 75 mg
intramuscular dose is equivalent to an oral dose of 88
mg, much greater than the 50 mg dose proven to be
effective in a prior study. However, the 50 mg oral
dose was slightly less effective in gestations more
than 42 days as compared with those up to 42 days
(88% vs 95%, p = 0.09).2 There is clearly no differ-
ence in effectiveness with the 75 mg intramuscular
dose based on gestational age with 95% efficacy for
both groups.
A potential concern with using a smaller dose of
methotrexate is a potential for lower immediate suc-
cess and continuing embryonic cardiac activity 1
week after methotrexate treatment. Because the pro-
tocol requires these patients to return 1 week after
the second dose of misoprostol rather than 4 weeks
later, this could create extra visits for patients,
thereby increasing cost and provider inconvenience to
see the patient more often. The number of subjects
requiring a day 15 visit in this study was 8.1%. In the
published report using methotrexate 50 mg/m2 intra-
muscularly, 32 of 299 women (10.7%) 5 56 days’
gestation required this extra visit.’ Review of the
original data shows a need for day 15 visits by 10 of
202 women (5.0%) up to 49 days’ gestation, a rate
similar to the 8.1% rate in this trial (p = 0.3 1).
Pain medication use and return of menses were
similar to those reported with regimens using 50
mg/m2 methotrexate intramuscularly’ and 50 mg
orally.2 Although the amount of methotrexate admin-
istered to all but three subjects was less with the 75
mg dose than with a calculated 50 mg/m2 dose,l the
reported rate of nausea (47% vs 19%), vomiting (12%
vs 9%), diarrhea (22% vs 7%), headache (16% vs 9%),
dizziness (21% vs 4%), and warmth/hot flashes (43%
vs 3%) after methotrexate administration was greater.
This finding demonstrates the difficulty with com-
paring side effects from two different trials and is
likely due to varying reporting by distinct patient
populations and ascertainment bias.
This study confirms that lower doses of methotrex-
ate, when administered intramuscularly, are equally
effective to a regimen based on a dose of 50 mg/m2.1
A caveat is that the activity of this dose against
undiagnosed extrauterine pregnancy is unknown. Ad-
ditionally, the lowest effective dose of methotrexate,
when combined with misoprostol for abortion, needs
to be determined.