Analysis of Severe Adverse Events Related to the Use of Mifepristone as an Abortifacient

Margaret M. Gary, MD; Donna J. Harrison, MD


The Annals of Pharmacotherapy. 2006;40(2):191-197. 

In This Article


The most frequent serious (CTCAE 3) and life-threatening (CTCAE 4) adverse events were hemorrhage (210 pts.), infection (46 pts.), and missed diagnosis of ectopic pregnancy (17 pts.). Table 2 contains the number of AERs by age and CTCAE grade. Five deaths (CTCAE 5) were reported,[11,12,13,14,15] including one 16-year-old,[11] along with 64 life-threatening events (CTCAE 4) and 224 serious events (CTCAE 3). The distribution of cases per CTCAE grade by age was roughly proportional except for deaths. The single death occurring in the 13-17 year age category represented 20% of the deaths, whereas this age group presented only 2% of all AERs (13 of 607).

Table 3 classifies AERs by CTCAE grade and diagnosis, including information on the group aged 13-17 years. Because the pediatric age encompasses a time of physical maturation, much of which is hormonally mediated, special scrutiny of this population is warranted. Both the US and French trial data specifically excluded women under 18 years of age; thus, the adverse events analyzed here present uncommon public information available on mifepristone's clinical use in the 13-17 year age range.

The 607 AERs included 5 deaths: 2 Californians, from sepsis;[13,14] a Tennessee woman with a ruptured ectopic pregnancy;[12] a Swedish teen, from massive hemorrhage;[11] and a British female, from "unknown etiology."[15] This last patient presented to the emergency department in shock and was found on autopsy to have 1 liter of blood in her stomach and 2 gastric ulcers. Sepsis is a known risk factor for stress-related gastrointestinal bleeding;[16] thus, sepsis is a plausible etiology for shock in this patient. Three deaths were not documented in these AERs: a participant in Canadian trials, from sepsis; an Asian Californian, from sepsis (December 2003); and a white Californian, from sepsis (June 2005). The FDA recently announced findings from the Centers for Disease Control and Prevention that all 5 of the sepsis deaths (4 Americans, 1 Canadian) have been linked to Clostridium sordellii.[17] Thus, there has been a total of 8 known deaths to date, including 5 Americans.

The severity of hemorrhage was coded based on the patient's pretransfusion hemoglobin level, the number of units transfused, the location and nature of surgical intervention, and the clinical description of the event according to the guidelines for CTCAE coding of hemorrhagic adverse events. However, in the majority of the AERs reporting hemorrhage, critical information (eg, baseline blood counts, vital signs, number of units transfused, posttransfusion hemoglobin or hematocrit level) was absent. Forty-two women experienced a life-threatening hemorrhage, as defined by active hemorrhaging with hemoglobin less than 7 g/dL and the transfusion of 2 or more units of packed red blood cells (PRBCs). One hundred sixty-eight women had severe hemorrhage, defined by hemoglobin of 7 g/dL or above and transfusion. Overall, 39% of AERs reported hemorrhage.

Serious or life-threatening infections were reported for at least 46 women, of whom 2 were aged 13-17 years.[18,19,20] Four women who had life-threatening infections but survived were in septic shock at the time of presentation to the emergency department.[18,19,21,22] One patient (aged 15 y) presented with adult respiratory distress syndrome (ARDS) from sepsis.[18] A second patient presented with ARDS from Escherichia coli sepsis.[21] A third presented with toxic shock syndrome.[22] A fourth (aged 16 y) presented with group B Streptococcus septicemia.[19] In addition to these 4 patients with documented infectious etiology, a fifth patient presented with disseminated intravascular coagulopathy (DIC) with hepatic and renal failure.[23] Her AER exhibited poor documentation, including only information on outpatient treatment for bacterial vaginosis. Because there was no documentation of infection, she was not included in the CTCAE 4 category for infection, although it is probable that the DIC was infectious in origin. Forty-three additional women required parenteral antibiotics for severe pelvic infection, and an additional 14 were treated for pelvic infections as outpatients. Overall, 11% of AERs reported an infectious complication.

The third most common class of severe adverse events was undiagnosed ectopic pregnancy. Seventeen patients had ectopic pregnancies that were undetected at the time of mifepristone administration. Eleven of these were ruptured at the time of diagnosis[12,24,25,26,27,28,29,30,31,32,33] (CTCAE genitourinary grade 4), including one death (CTCAE 5).[12] Three of the 17 were cornual pregnancies,[31,33,34] and one was a cervical pregnancy that resulted in a hysterectomy in a 29-year-old patient.[35] Overall, 2.8% of AERs reported unrecognized ectopic pregnancy.

Seven additional diagnoses of possible sentinel events were (1) myocardial infarction in a previously healthy 21-year-old woman,[36] (2) prolongation of QT interval on electrocardiogram in another woman,[37] (3) pulmonary embolism,[38] (4) exacerbation of Crohn's disease,[39] (5) precipitation of a sickle cell crisis,[40] (6) acute pancreatitis,[41] and (7) drug interaction resulting in liver failure in an HIV-positive patient.[42]

One unexpected finding was a number of allergic reactions ranging from hives to severe generalized urticaria. One patient was hospitalized for 4 days and treated with intravenous diphenhydramine and oral prednisone.[43] Eight patients were treated as outpatients, and 4 required no treatment.

The extent of treatment required to manage the adverse events is presented in Table 4 . Sixty-eight women received transfusions. Nineteen (28%) of these required 3 or more units of PRBCs. In 15% of cases with transfusion, the number of units was not documented in the AER. At least 513 surgical procedures were performed in the 607 patients with adverse events. When a patient had more than one surgery performed, only the most extensive surgery was included in the tabulation. (For example, if a patient had a dilatation and curettage and then a subsequent laparotomy, only the laparotomy was included. For patients with more than one dilatation and curettage, only one was included in the tabulation.) There were 235 emergency surgeries performed. Of these, 17 (7%) were emergency laparotomies: 16 were for ectopic pregnancies (1 ectopic pregnancy was managed laparoscopically) and one laparotomy was for sepsis. Two of the 5 deaths were intraoperative. The remaining 93% of emergency surgeries were emergency dilatation and curettage procedures performed to arrest hemorrhage. At least 40% of the patients were hospitalized for treatment, including 12 admissions to the intensive care unit. Fifty-seven percent were managed as outpatients and, in 3%, the site of treatment was not documented.

Because mifepristone abortions result in exposure of the fetus to known teratogens, the AERs were analyzed for information on fetal outcome after exposure, where documented ( Table 5 ). Of the 278 pregnancies terminated after mifepristone failure without other diagnoses, 58 (21%) had documented fetal viability by ultrasound on return visit, and 36 (13%) had documented fetal demise or retained products of conception without a fetus. However, 184 (66%) had no documentation of fetal status. Of the 22 fetuses documented to be viable in the second trimester, 9 were lost to follow-up and 13 had a documented outcome to the pregnancy. Of those 13, 9 had a termination of pregnancy with no documentation of fetal morphology, 1 was enrolled in a fetal registry,[44] and 3 had documentation of fetal anomalies: (1) Mobius syndrome,[45] (2) neural tube defect,[46] and (3) oligodactylia, monodactylia, facial dysmorphia, and meningo-encephalocele.[47] (One additional AER[48] referred to a separate fetal report implying malformation. However, that fetal report was not included in the information released and thus not counted in our calculation of malformation rate.) Assuming normal fetal morphology in all 9 pregnancies that were terminated without documenting information about the presence or absence of malformations and assuming normal fetal morphology for the patient enrolled in the fetal registry, the rate of fetal malformations would still be at least 3 (23%) of 13.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: