methotrexate and misoprostol for early abortion: a multicenter trial. acceptability

Methotrexate and misoprostol for early abortion

Mitchell D. Creinin, M.D., and Philip D. Darney, M.D., M.Sc.

Department of Obstetrics, Gynecology and Reproductive Sciences, University of California, San Francisco, San Francisco General Hospital

Methotrexate is cytotoxic to trophoblast and, in low doses, has minimal side effects. It is used to treat both gestational trophoblastic neopla- sia and ectopic pregnancy. The cytotoxic effects of methotrexate on intrauterine trophoblast should be equivalent. To test this hypothesis, ten pregnant women, ~8 weeks’ gestation were reated with methotrexate 50 mglm” intramuscularly followed 3 days later by misoprostol, a prostaglandin E, analogue. The first 4 patients received misoprostol 600 pg orally; none aborted soon after the misoprostol. Two patients aborted 25 and 26 days after the methotrexate injection and two elected a suction abortion after 14 days (one by choice and one because the pregnancy was still viable). The last 6 patients received misoprostol800 pg vaginally and aborted within 3-8 hours. One patient had an incomplete abortion requiring a suction curettage 34 days after the misoprostol. Vaginal bleeding for these 6 patients lasted an average of 29 ? 11 days (range, 12-42 days). No methotrexate side effects were observed. Vaginal misoprostol (800 pg) was significantly more effective (p = 0.005) than oral misoprostol (600 kg) in effecting abortion after intramuscular methotrexate.

Keywords: Medical abortion; methotrexate; misoprostol; P-hCG; vaginal ultrasound.

Introduction

Medical abortion has become available in Europe and China as an alterna- tive to a surgical abortion. A medical procedure, however, has two very important and unique benefits: 1) it allows a woman to have a safe, effective abortion without an invasive surgical procedure; and 2) it could

Submitted for publication June 25, 1993; accepted for publication July 25, 1993.

0 1993 Butterworth-Heinemann Contraception 1993:48, October 339

Methotrexatelmisoprostol abortion: Creinin and Darney

potentially make safe abortion more accessible. Surgical abortion is not available in 80% of counties in the United States. Abortion is even less available in developing countries where lOO,OOO-200,000 women die yearly from complications of aborti0n.l

Methotrexate has been used in the treatment of molar pregnancies and choriocarcinoma for decades. In 1982, Tanaka and co-workers successfully used methotrexate to treat a patient with an ectopic pregnancy.2 Since then, more than 25 reports totaling more than 400 subjects have described the use of methotrexate as a treatment for ectopic pregnancy in selected patients. Follow-up studies of these patients have demonstrated that methotrexate does not impair future fertility, menses returns normally, and pregnancy rates are similar to those achieved by traditional surgical treatment of ectopic pregnancy.,7 Additionally, the peak serum level of methotrexate after intramuscular injection for ectopic pregnancy has been shown to be significantly lower than the accepted toxic level.4

In 1991, Stovall and colleagues reported the successful use of methotrexate for ectopic pregnancy in a single-dose treatment without leuco- vorin rescue.” This regimen, consisting of methotrexate 50 mg/m2 injected intramuscularly, has the benefits of single-dose treatment, high efficacy, low morbidity and low cost. Stovall and colleagues recently updated their series of cases, reporting on 121 patients treated with this methotrexate regimen with a 94.3% success rate and minimal side effects.”

The cytotoxic effects of methotrexate on intrauterine trophoblastic tissue in early pregnancy should be equivalent. A recent case report described the safety and efficacy of methotrexate injected directly into the amniotic fluid and placenta (25 mg at each site) to cause abortion of an 8 week intrauterine pregnancy.’ Misoprostol has been safely used in Brit- ain and France” with RU-486 (mifepristone) for abortion. Thus, we performed a pilot study to assess the efficacy of methotrexate in combination with misoprostol to induce abortion for women <8 weeks’ gestation.

Methods

Ten patients were included in a prospective trial approved by the Institu- tional Review Board of the University of California, San Francisco. Healthy English- and Spanish-speaking women were recruited from the Family Planning Clinic at San Francisco General Hospital and the Gyne- cology Clinic at Mt. Zion Hospital, San Francisco, California. Entry criteria included: 1) 2 18 years of age, 2) requesting an elective abortion, 3) with an intrauterine pregnancy of <8 weeks’ gestation documented by vaginal ultrasound, 4) willing to abstain from intercourse for the first 21 days of the study and comply with the visit schedule, 5) with adequate venous access for multiple phlebotomies, and 6) with access to a tele- phone.

340 Contraception 1993:48, October


Exclusion criteria included: 1) use of prenatal vitamins or any other medications containing folate, 2) hematocrit c30%, 3) white blood cell count ~3.0 thousand/fl, 4) platelet count <lOO thousand/fl, 5) AST >2X normal or active liver disease, 6) serum creatinine >1.5 mg/dl or active renal disease, 7) inflammatory bowel disease, and 8) known intolerance of or allergy to methotrexate or misoprostol.

The study protocol, risks, benefits, visit schedule and consents were reviewed with each potential subject by the first author. Each patient who enrolled signed an informed consent and agreed to suction abortion should the pregnancy still be viable (defined as the presence of cardiac activity on vaginal ultrasound) 14 days after initiating the study. Vaginal ultrasound (General Electric RT 3200, Milwaukee, WI) was performed and the pregnancy considered normal if it was intrauterine, a yolk sac was present, and a fetal pole with cardiac activity was present. Gestational age was based on last menstrual period (LMP) but was changed if the age based on the initial crown-rump length differed from the gestational age by LMP by more than 4 days.‘O

After obtaining a baseline complete blood count, platelet count, AST, serum creatinine, blood type (and screen if Rh-negative), and quantitative serum B-hCG, each patient received methotrexate 50 mg/m2 intramuscularly. The day of the injection was considered study day 1. If the patient’s blood type was Rh-negative, she received Rh-immune globulin intramuscularly on day 1. The patient returned to the clinic each day for the next 6 days; a vaginal ultrasound was performed and serum B-hCG measured. Normal B-hCG doubling time was defined as an increase >66% over 48 hours. Serum B-hCG was measured by immunoenzymatic assay (Tosoh Medics AIA-600, South San Francisco. CA) in the second International standard.

All patients received misoprostol when they returned to the clinic on study day 4. The patients then stayed in the clinic for 2-3 hours after receiving the misoprostol to be observed for side effects. The first four patients received misoprostol 600 pg orally. To assure compliance, the patients swallowed the tablets under supervision. Because this dose proved ineffective, the regimen was changed to misoprostol800 pg vaginally to be repeated in 24 hours if the pregnancy was still evident on vaginal ultrasound; this change in the protocol was approved by the Insti- tutional Review Board. The misoprostol was administered through a speculum after the vaginal ultrasound was performed; the gel remaining from the ultrasound was wiped from the vagina, four 200 pg tablets were placed into the vagina through the speculum, and a large cotton swab was then used to push the tablets into the posterior fornix as the speculum was removed. The patients were allowed to sit or lie down, at their own discretion, after administration of the misoprostol.

Each patient was asked to keep a symptom log of vaginal bleeding,

Contraception 1993:48, October 341


passage of tissue, abdominal cramping, headache, dizziness, nausea, vomiting, and oral ulcers. On day 7 or 14, a complete blood count, platelet count and AST were obtained. The serum @hCG was followed weekly until c 10 mIU/ml.

Outcome measures assessed included successful abortion, cessation of fetal cardiac activity, onset/duration of vaginal bleeding, incidence of reported side effects, and change in hemoglobin, AST, and serum P-hCG. The trial was analyzed as a cohort study by comparing the experience of those treated with misoprostol orally vs. vaginally. Notwithstanding the cohort analysis, details of individual cases are presented because of the novelty of the treatment method. Fisher exact test or Mann-Whitney U test (for non-parametric data) was performed as appropriate with a p ~0.05 considered significant. Serum @hCG values were compared after logarith- mic transformation.

Results

Patients in both groups were similar (Table 1). Patients who received misoprostol orally were not significantly different than patients who received misoprostol vaginally in regard to age, gravidy, parity, body surface area, gestational age, or initial @hCG value. None of the patients had vaginal bleeding or abdominal cramping prior to entry into the study. One patient in the oral misoprostol group and three in the vaginal misoprostol group had a prior surgical abortion. The serum p-hCG did not increase at a normal rate within the first 48 hours after the methotrexate injection in all patients.

None of the first four patients aborted after misoprostol(600 ug orally) on study day 4, although patients 1 and 2 did have vaginal bleeding. Patient 1 had light vaginal bleeding beginning 3 hours after receiving the misoprostol and stopping after 24 hours. Patient 2 had menses-like bleeding beginning 4 hours after receiving the misoprostol; 36 hours later the bleeding decreased in intensity and she continued to spot for 8 more days. Serum P-hCG values for the first 4 patients are shown in Figure 1.

Table 1. Selected characteristics of patients receiving methotrexate followed by oral vs. vaginal misoprostol

(mean + standard deviation).

Misoprostol Age Gravidy Parity

Body Gestational Surface Age (days)

Area (m2) Day 1 p-hCG Day 1

oral 600 pg (n = 4) 32.8 (5.8) 3.3 (1.9) 1.0 (2.0) 1.64 (0.23) 46.8 (2.1) 51,463 (24,088) vaginal 800 pg (n = 6) 29.7 (6.6) 3.3 (1.2) 1.7 (1 .O) 1.63 (0.13) 48.7 (2.9) 74,833 (36,411)

Contraception 1993:48, October


”

Patient 1 had cessation of fetal cardiac activity by day 2 and declining P-hCG values. However, when the pregnancy was still not expelled by day 14, she elected to have a surgical abortion which was performed without complication that same day. Patient 4, despite declining P-hCG levels, continued to have fetal cardiac activity on day 14. The fetal crown- rump length, though, had only increased from 7 mm (49 days’ gestation) on day 1 to 15 mm (57 days’ gestation) on day 14. Because cardiac activity was present, even though fetal growth was less than normal and the f3- hCG was less than the level on day 1, the patient underwent a surgical abortion on day 1.5.

Patients 2 and 3 had a fetal pole without cardiac activity and declining P-hCG values on day 14 and elected to wait for spontaneous expulsion. Both patients were followed with a weekly vaginal ultrasound examina-

Study Day

FIGURE 1. Serum p-hCG values for patients l-4 who received methotrexate followed by misoprostol 600 Fg orally, The p-hCG increased from day 1 to 7 by 27.1%, 11.3%, 17.1%, and 3.1%, respectively



tion and serum P-hCG. Patient 2 began to spot again on day 24 followed by heavy menstrual-like bleeding and cramping on day 26. Vaginal ultrasound on day 29 showed an empty uterus with a normal endometrial stripe; the hemoglobin that same day was equivalent to the baseline value. The bleeding slowed on day 30 and continued as light vaginal bleeding until day 34. She spotted for 2 more days, then all bleeding stopped.

Patient 3 experienced no effects from either medication until she began spotting on day 25 followed by heavy vaginal bleeding later that evening. The bleeding decreased by the next morning to normal menstrual-like flow and continued until day 33. The patient denied any pelvic cramping with the bleeding.

Abortion occurred in 100% (9.5% CI 54-100%) of the next six patients within 3-8 hours after vaginal misoprostol administration; none required a repeat dose. Vaginal ultrasound examination the following day showed the uterus to contain a scant amount of blood in the endometrial cavity in all patients. Vaginal bleeding lasted an average of 29 + 11 days (range 12- 42 days). Two patients felt the bleeding was heavier than a normal period. The bleeding was similar to menstrual bleeding within 24 hours and no greater than “spotting” by the tenth day.

Vaginally-administered misoprostol was significantly more effective than orally-administered misoprostol at effecting abortion after intramuscular methotrexate. All six patients who received misoprostol800 kg vaginally aborted within two weeks compared to none of four who received 600 pg orally (p = 0.005). Complete abortion occurred in 5/6 patients (83%, 95 % C136-lOO%)andtheirserump-hCGvalues (Figure2)decreasedto<lOmIU/ ml by day 21-50 (average 36 + 9 days) One patient required suction curettage on day38 for an incomplete abortion. She presented with heavy vaginal bleeding and an open cervical OS; pathology revealed necrotic villi and de- cidua. Serum P-hCG on that day was 30 mIU/ml.

Fetal cardiac activity was no longer present on vaginal ultrasound on day 2 for the first patient, day 3 for the second patient, day 5 for the third patient, and day 3 for the tenth patient. All other patients who received misoprostol vaginally expelled the pregnancy before cardiac activity ceased; however, 4 of the 5 fetuses had slow cardiac activity (<I20 beats per minute) on the day of misoprostol administration.

Side effects were minimal. The hemoglobin, platelet count and AST showed no appreciable change from baseline when evaluated on day 7 or 14. Symptom logs were completed by 9 of the 10 patients. After the methotrexate, 4 patients experienced mild abdominal cramping on the first or second day. Six of the patients experienced headache, dizziness, nausea, and/or vomiting related to the pregnancy. In five of these women, all of whom received misoprostol vaginally, these symptoms were relieved after the pregnancy was expelled a few hours after misoprostol administration; the sixth (patient 4) continued to feel “morning sickness” through



\__

I

i. . _’ .

k\ 1::

\

1 ‘.

. . . . . ‘1:

\

\

ll- 7

\ : \ B \ \

-

\ ‘4

‘..

\

7- .

\

. .

“. . . . .

‘\ ‘.

\

‘. L

\

I-

. .

” I..

.

‘N.

f' 8

l

f- .

. . . .

.

,7-

_.-. __.- -..- --- . . . . . . . . . . . .

•l 5

. 6

7

,’ 6

9

- 10

-..... . . .

1-m-

‘. . . . .

T

“.. . . .

-

49;

FIGURE 2. Serum P-hCG values for patients 5-10 who received methotrexate followed by misoprostol 800 pg vaginally. The P-hCG increased from day 1 to 7 by 44.1%, 14.3%, 0, 3.6%, 27.5%, and 9.7%, respectively

day 9. One patient who received the misoprostol orally and one patient (2 days after misoprostol administration) who received the misoprostol vaginally reported nausea. No patient experienced vomiting after misoprostol administration.

Five patients experienced abdominal cramping on the day of misopros- to1 administration (2 oral, 3 vaginal); one patient who received misoprostol vaginally continued to have mild cramping through day 13. All cramping was easily relieved with ibuprofen 600 mg orally. Diarrhea occurred in 4 patients after the misoprostol (2 oral, 2 vaginal); the diarrhea was mild and limited, occurring 2-4 times over a 12-hour period.

Discussion

Methotrexate and vaginally administered misoprostol can be an effective abortifacient for pregnancies ~8 weeks’ gestation. All of the patients who



received methotrexate followed by misoprostol800 pg vaginally had rapid expulsion of the intrauterine products, and 5/6 (83%) had a complete abortion. Vaginally-administered misoprostol was significantly more effective than was orally administered misoprostol, but a larger dose was used (800 pg vs. 600 kg).

All pregnancies showed effects of methotrexate-induced trophoblast cytotoxicity as evidenced by the abnormal rate of P-hCG increase and rapid embryonic death in patients 1,2,3, and 10. The effect of methotrexate on the amount of @hCG increase over 48 hours was variable and did not appear related to gestational age or initial P-hCG level. All of the pregnancies were “normal” prior to treatment as defined by the presence of cardiac activity on vaginal ultrasound and the lack of vaginal bleeding prior to entry into the study. Pregnancies meeting these qualifications have a spontaneous loss rate of approximately 5% (Goldstein S, personal communication, June 2, 1993). More specifically, the spontaneous loss rate with a crown-rump length <5 mm is 7.2% and 5-10 mm, 3.3%.”

Vaginal bleeding was prolonged, but was not heavier than “spotting” after the tenth day. The length of bleeding may reflect the manner in which methotrexate acts to disrupt the supporting trophoblast or the ability of the misoprostol to effect complete expulsion at various gestational ages.

Side effects from methotrexate were limited to mild abdominal cramping in 3 patients. Gastrointestinal side effects were minor and occurred more often with oral administration of misoprostol as compared to vaginal. Only 1 of 5 patients who received misoprostol vaginally and completed a symptom log reported any gastrointestinal side effects (diarrhea). In a recent French report with RU-486 600 mg followed 2 days later with misoprostol 400 pg orally, the rates of nausea, vomiting, and diarrhea following the oral misoprostol were 43%, 17%, and 14%, respectively.” The incidence of gastrointestinal side effects with misoprostol appears related to the route of administration; an 800 kg dose vaginally had minimal side effects as compared to 400 pg orally.

The number of patients treated is too small to make comparisons to RU-486. Unlike RU-486, methotrexate is an antimetabolite which could damage a fetus carried to term after intrauterine exposure. Reports of teratogenicity involve high doses of methotrexate as used for chemotherapy or exceeding normal dose ranges.” With low-doses, reports are scant although a review of teratogenicity with low-dose oral methotrexate in early pregnancy found no effect.‘,’ Teratogenicity is a concern since 6 of the 10 pregnancies in this report had ultrasound-detectable cardiac activity at the time of misoprostol administration; still, concern that patients may not return for the misoprostol is minimized by evidence from French and British RU-486 trials where all patients returned for the prostaglandin.la In addition, future fertility and congenital anomalies in subsequent preg-



nancies are not effected by methotrexate as evidenced by its use in the treatment of gestational trophoblastic neop1asia.1~

Additionally, methotrexate is widely available in the United States and inexpensive. Two vials of methotrexate [ 50 mg each), which would provide an adequate treatment dose for the majority of patients, costs $3.78 at our institution. Misoprostol is also inexpensive, costing $0.54 for four 200 f_~g tablets. Bleeding with this regimen, though, appears to last longer than with RU-486 and misoprostol.’ While the comparative efficacy and safety of these two approaches are unknown, methotrexate plus vaginal misoprostol offers the prospect of medically-induced abortion with cur- rently available and inexpensive drugs. Although this makes methotrexate and misoprostol treatment for abortion presently attractive, the use of this regimen is investigational.

References

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

Darney PD. Maternal deaths in the less developed world: preventable trage- dies. Int J Gynaecol Obstet 1988;26:177-9. Tanaka T, Hayashi H, Kutsuzawa T et al. Treatment of interstitial ectopic pregnancy with methotrexate: report of a successful case. Fertil Steril 1982;37:851-2. Stovall TG, Ling FW, Buster JE. Reproductive performance after methotrexate treatment of ectopic pregnancy. Am J Obstet Gynecol 1990;162:1620-4. Schiff E, Shalev E, Bustan M et al. Pharmacokinetics of methotrexate after local tubal injection for conservative treatment of ectopic pregnancy. Fertil Steril 1992;57:688-90. Stovall TG, Ling FW, Gray LA. Single-dose methotrexate for treatment of ectopic pregnancy. Obstet Gynecol 1991;77:754-7. Stovall TG, Ling FW. Single-dose methotrexate: an expanded clinical trial. Am J Obstet Gynecol 1993;168:1759-65. Buckshee K, Dhond AJ. A new nonsurgical technique for termination of intrauterine pregnancy associated with large multiple uterine leiomyomas. Int J Gynaecol Obstet 1992;37:297-9. Norman JE, Thong KJ, Baird DT. Uterine contractility and induction of abortion in early pregnancy by misoprostol and mifepristone. Lancet ii 1991;338:1233-6. Peyron R, Aubeny E, Targosz V et al. Early termination of pregnancy with mifepristone (RU 486) and the orally active prostaglandin misoprostol. N Engl J Med 1993;328:1509-13. Goldstein S. Embryonic ultrasonographic measurements: crown rump length revisited. Am J Obstet Gynecol 1991;165:497-50. Hill LM, Guzick D, Fries F et al. Fetal loss rate after ultrasonically documented cardiac activity between 6 and 14 weeks menstrual age. J Clin Ultra- sound 1991;19:221-3. Darab DJ, Minkoff R, Sciote J, Sulik KK. Pathogenesis of median facial clefts in mice treated with methotrexate. Teratology 1987;36:77-86. Kozlowski RD, Steinbrunner JV, MacKenzie AH et al. Outcome of first-



trimester exposure to low-dose methotrexate in eight patients with rheu- matic disease. Am J Med 1990;88:589-92.

14. Lader L. RU 486. The pill that could end the abortion wars and why American women don’t have it. Menlo Park, CA: Addison-Wesley Publishing Company, Inc., 199158-9.

15. Ross GT. Congenital anomalies among children born of mothers receiving chemotherapy for gestational trophoblastic neoplasms. Cancer 1976;37: 1043-7.

348 Contraception l993:48, October

methotrexate and misoprostol for early abortion: a multicenter trial. acceptability

Documents