Controlling refractory atonic postpartum hemorrhage with Hemabate sterile solution

Controlling refractory atonic postpartum hemorrhage with Hemabate sterile solution MerriKay A. Oleen, PhD, and Joseph P. Mariano, MD Kalamazoo, Michigan The customary use of Hemabate sterile solution for postpartum hemorrhage was studied at 12 cooperating obstetrics units for a 12-month penod. Outcomes of interest were the characterislics of patients chosen by the attending physicians to receive the drug, conditions of drug use, and patient status after drug use. Cessation of bleeding was considered a successful outcome and in 208 of 237 cases (87.8%) the hemorrhage was controlled directly after the administration of Hemabate sterile solution. An additional 17 successful outcomes were achieved with further administration of oxytocics for an overall success rate of 94.9%. Twelve cases of postpartum hemorrhage were considered pharmacologic treatment failures, requiring surgical intervention. Among the patients in whom pharmacologic treatment failed were factors that may have played a significant role in the cause of the hemorrhage including peripheral coagulopathy, retained products of conception, lacerations, chorioamnionitis, oxytocin-induced or augmented labor, increased fetal weight, magnesium-treated preeclampsia, and cesarean delivery. However, no combination of factors could be consistently associated with pharmacologic treatment failure. (AM J OSSTET GVNECOL 1990;162:205-8.)

Key words: Postpartum hemorrhage, drug treatment, risk factors

The leading cause of early postpartum hemorrhage is uterine atony. The commonly accepted management scheme for the control of hemorrhage as a result of uterine atony includes the use of manual massage and bimanual compression supplemented promptlY by restoration of blood volume and administration of oX\tocics. If initial treatment with an oxytocin infusion yields no response an intramuscular or an intravenous injection of ergot alkaloid derivatives mav be given. I When the physical maneuvers and pharmacologic therapies fail, surgical techniques must be used. Ligation of the arterial supply of the uterus, the uterine or hypogastric arteries, mav succeed in stopping atonic postpartum hemorrhage, but when conservative surgery will not control bleeding, removal of the uterus bv hysterectomv is necessary." During the past several years mvestIgators hdve sought alternative treatment modalities to hysterectomy for severe atonic postpartum hemorrhage, and the most favorable and encouraging results have been seen with some of the prostaglandins. They appear to be involved in postpartum hemostatic mechanisms by virtue of their pharmacodynamic properties relative to

From The Upjohn Compan.~. A complete Itst of partlClpattng mstltutlOns IS given at the end of the artIcle. Hemahate ,terlle solutIOn IS a reg1Stered trademark of The UpJohn Company. Recewedfor puhhcation March 23,1989; reVISed May 31.1989, accepted June 22, 1989. Reprmt requests: MerrzKa.~ A Oleen. PhD, Drug EpIdemIOlogy Unzt, 9141-243-135, The UP1"hn Company, Kalamazoo, M149001.

611 114832

myometrial stimulation, vasoactive effects, and platelet function."' The compound that has been approved by the U.S. Food and Drug Administration for the treatment of atonic postpartum hemorrhage refractory to conventional methods of management is (l5S)-15methyl prostaglandin F 2n -tromethamine (Hemabate sterile solution, formerly Prostinl 15M sterile solution, The Upjohn Company, Kalamazoo, Mich.). When the U.S. Food and Drug Administration granted approval for the postpartum hemorrhage indication for this product they requested a proposal for an epidemiologic investigation of the customary use of Hemabate sterile solution during the initial 12 months of market availability for the new indication. The purpose of this article is to report the results of the multicenter surveillance that was designed to describe the patients chosen by the attending physicians to receive Hemabate sterile solution for the treatment of postpartum hemorrhage, the conditions of clinical use in this indication, and patient outcomes.

Material and methods A prospective epidemiologic study design that uses a method of data collection by obstetrics staff at multiple centers was adopted. The Medication Monitoring Program Coordinating Center recruited 12 obstetrics units at nonprofit and governmental hospitals that agreed to participate in this study. They were selected on the basis of annual estimated cases of postpartum hemorrhage and capability to provide medical record information concurrently with hospitalization for all

205

206 Oleen and Mariano

January 1990 Am J Obstet Gynecol

Table I. Conditions of delivery in patients with postpartum hemorrhage (n = 237) Condztzons

Frequency

% of Cases

Oxytocin-induced or augmented labor Cesarean delivery Magnesium-treated preeclampsia Lacerations Retained products of conception Increased fetal weight Peripheral coagulopathy Chorioamnionitis

92

3S.S

72 43

30.4 1S.1

35 27 22 4 3

14.S 11.4 9.3 1.7 1.3

Table II. Associated pathology among patients in whom drug treatment failed (n = 12) A rtery ligation (n = 5)

Laceration, 2 Placenta previa, 1 Hypertension placenta accreta (partial) , 1

Artery ligation and hysterectomy (n = 4)

Retained products of conception, 1 Retained products of conceptionplacenta previa-placenta accreta, 1 Chorioamnionitis, 1

Hysterectomy (n = 3)

Peripheral coagulopathyplacental abruption, 1

patients selected by their physicians to receive Hemabate sterile solution in a 12-month period between January 1986 and March 1987. In all cases the data were abstracted from the medical charts by attending nurses and physicians during and immediately after the patients' hospitalizations. A specially designed case report form was completed for each patient to record demographic characteristics; pregnancy history; a clinical report of labor and delivery for the current admission and clinical events associated with the postpartum hemorrhage, whether acute or delayed, including estimated blood loss and the mechanical and pharmacologic therapies applied. The Hemabate sterile solution exposure was characterized by documented dosage administered and its ranked order of administration relative to all pharmacologic therapies used in the postpartum hemorrhage management scheme. Information concerning the presence of genital tract trauma, retained products of conception, and peripheral coagulopathy were also collected because the failure of Hemabate sterile solution to control refractory hemorrhage could be related to these pathologies in view of its mechanism of action.

Patient outcomes to be recorded were the cessation of bleeding and, thus, the lack of need for artery ligation or emergency hysterectomy and any medical events associated with Hemabate sterile solution treatment. The person who completed the case report form was asked to note if a charted medical event was, on the basis of a physician's or nurse's medical judgment, considered to be drug-related . Other factors known to influence the outcome of atonic postpartum hemorrhage such as the presence of chorioamnionitis, oxytocin augmentation of labor, and increased fetal weight were also included on the case report form to be appropriately controlled in the analysis of the study data. The case report forms were returned to the Coordinating Center on a monthly schedule and the information was entered in the data base. Analysis consisted of descriptive statistics to characterize the patients selected to receive Hemabate sterile solution for postpartum hemorrhage, the customary conditions for the drug'S use in this indication, and the patient outcomes. Further subgroup analyses were done to look for differences among the cases on the basis of the presence of known risk factors for poor outcome in postpartum hemorrhage. Results

The 236 study participants had 237 episodes of postpartum hemorrhage. The mean age ofthe patients was 25.3 ± 5.7 years. There were 113 multiparous women, 108 primiparous women; and 15 women who had not borne viable offspring. The breakdown of the method of delivery was 72 cesarean sections and 164 spontaneous vaginal deliveries. Table I lists conditions associated with postpartum hemorrhage that were recorded for patients. There were reports of associated risk factors for poor outcome at delivery for 174 cases, 84 of them with multiple factors. Hemorrhage was successfully controlled immediately after the administration of Hemabate sterile solution in 208 of the 237 cases (87.8 %). An additional 17 successful outcomes were achieved with further administration of oxytocics for an overall success rate of94.9%. Included as instances in which the therapy failed were patients for whom surgery was performed after administration of Hemabate sterile solution. Five patients required artery ligation, four patients underwent artery ligation that preceded hysterectomy, and an additional three patients required immediate hysterectomy. Table II presents the associated pathology among the instances of drug treatment failure . The conventional management scheme to control postpartum hemorrhage was followed in nearly all cases ; 96.2% of the patients received oxytocics before Hemabate sterile solution. The mean number of doses

Volume 162 Number 1

(125 or 250 J,Lg) of Hemabate sterile solution given to these patients including those in whom therapy failed, was 1.16. The range was one to five doses. There were 230 of the 237 patients (97%) who received only one or two doses. The intervals of time between multiple doses were determined by the attending physicians as dictated by the clinical events of the cases. The route of administration of Hemabate sterile solution was not systematically collected; however, there were voluntary reports that suggested the drug was administered deep, intramuscularly, as expected, and by various other routes including intramyometrially, intravenously, intrauterinely, and intracervicaUy. The blood losses were estimated before the administration of Hemabate sterile solution. Table III summarizes the estimates of blood loss for those who were successfully controlled and those for whom pharmacologic treatment failed. Only two cases of delayed postpartum hemorrhage were reported; one of these patients had an episode of postpartum hemorrhage with an estimated blood loss of 590 ml immediately after delivery and a recurrence 9 days later with an estimated 1000 ml blood loss. The other case had no prior history of postpartum hemorrhage; refractory hemorrhage developed 8 days after delivery. Only a portion of her blood loss was witnessed and quantified. Among the acute and delayed cases that were successfully treated the mean blood loss was 900 ± 748 ml; 57 of the cases had estimated blood losses of ::;500 ml. In contrast, the 12 patients for whom pharmacologic treatment failed had an estimated mean blood loss of 2229 ± 2454 ml and only two of these patients had estimated blood losses of < 1000 ml. This points to the severity of the hemorrhage among the failures. A total of 64 patients in this series (27.0%) received red blood cell-containing units with a mean of 3.6 units; the range was 1 to 29 units. Nine patients also received non-red blood cell-containing units such as fresh frozen plasma, cryoprecipitate, or albumin. A total of 48 of the 237 study participants had one or more side effects that the physician- or nurseabstractors attributed to therapy with Hemabate sterile solution (Table IV). The medical events that were included on the case report form were reported as adverse effects of varying frequencies related to the drug in clinical trials and published literature. There were 35 of the 48 patients who had only one drug-associated event. The most commonly reported medical events associated with the administration of Hemabate sterile solution were diarrhea (11.4% of237 patients), elevated blood pressure (6.8%), and vomiting (6.8%). None of the reports of elevated blood pressure was from cases in which hypertension was indicated to exist during the delivery. However, with the use of the recorded readings for the highest diastolic blood pressure in the first

Postpartum hemorrhage and Hemabate sterile solution

207

Table III. Estimated blood loss (ml) Total group (n = 227)*

successes (n = 215)

Therapy

Therapy fazlures (n = 12)

100-9500 970 ± 955

100-9500 900 ± 748

500-9500 2229 ± 2454

Range Mean

*Blood loss could not be estimated for 10 cases.

Table IV. Drug-related medical events Event

Frequency

% ofCases*

Diarrhea Elevated blood pressuret Vomiting Elevated temperature Flushing Tachycardia

27

11.4

16 16

5 4 4

6.8 6.8 2.1 1.7 1.7

*Percentages are calculated on the basis of237 treated cases. Some patients had more than one event. tDefined as >5 mm Hg diastolic over preparturition.

24 hours after the initial injection of Hemabate sterile solution as a guide to the intensity of the elevated blood pressure, it was noted that only 10 of the 16 reported cases had readings of 2:90 mm Hg and only 1 patient had a postpartum diastolic blood pressure reading as high as 110 mm Hg. Comment

These results support the conclusions of previous authors that the administration of Hemabate sterile solution is an effective and safe form of therapy for postpartum hemorrhage that is unresponsive to conventional nonsurgical therapy."6 Although the number of cases of postpartum hemorrhage as a result of uterine atony cannot be directly determined in this study, successful outcomes in terms of control of the hemorrhage were achieved directly after the administration of Hemabate sterile solution in 208 (87.8%) cases. Successful outcomes were achieved for an additional 17 patients with further administration of oxytocics after the Hemabate sterile solution for an overall success rate of 94.9%. This compares favorably with the success rates published by Toppozada et aU (93.8%), Hayashi et aI.' (86%), Buttino and Garite 6 (84.6%), and the rate of success reported in the clinical trials conducted in support of the postpartum hemorrhage indication collectively (87.8%).7 As in the clinical trials, there were factors in this study that may have played a significant role in the cause of the hemorrhage in the cases in which Hemabate sterile solution was judged to be ineffective. Among these fac-

208 Oleen and Mariano

January 1990 Am J Obstet Gynecol

tors were peripheral coagulopathy, retained products of conception, and lacerations present in five of the 12 pharmacologic treatment failures in this study. There is no documented efficacy of Hemabate sterile solution in cases of hemorrhage associated with these pathologies; when such therapy fails the patients must be examined closely for other causes of hemorrhage. Clark et al. 8 reported that in a case series of emergency hysterectomies for obstetric hemorrhage those performed for atony had a significant association with the following factors: (1) amnionitis, (2) cesarean section because of labor arrest, (3) oxytocin augmentation of labor, (4) magnesium sulfate infusion, and (5) increased fetal weight. Among the pharmacologic treatment failures in this study, 10 of the 12 patients involved were first seen with one or more of these risk factors for poor outcome. However, no combination of factors could be consistently associated with pharmacologic treatment failures. There were differences in both the range and mean estimated blood loss between the groups who either did or did not respond to Hemabate sterile solution therapy in this study as shown in Table III. Hayashi et al. 5 reported similar figures. Of interest were the 57 (24.1 %) successfully treated patients in this study with estimated blood losses of :5500 ml; these cases were reported from nine different obstetrics units. It is generally accepted, though arbitrarily defined, that postpartum hemorrhage exists when the estimated blood loss is >500 ml. 9 However, this study was designed to observe the customary use of drug treatment for the new indication, and no definition for postpartum hemorrhage was specified in the protocol. It appears that a clinical decision was involved in each of these 57 cases in which estimated blood loss was <500 ml, which led to the use of Hemabate sterile solution. Inasmuch as blood loss and replacement can be associated with serious complications, any mechanism to minimize risk including early use of Hemabate sterile solution appears to be justified. A total of 48 of the 237 patients (20.3%) in this study exhibited drug-associated medical events. The profile of reported events was similar to the medical events reported by Hayashi et al. 1. 5 and Toppozada et al. The gastrointestinal events, temperature elevation, and increased blood pressure level were also reported to be the drug-related medical events with greatest frequency OJ

in the clinical trials that supported the postpartum hemorrhage indication, collectively. 7 The incidence of drugassociated medical events reported in this study lends support to the safety and tolerance of Hemabate sterile solution in the treatment of postpartum hemorrhage. In summary, parenteral Hemabate sterile solution appears very effective for the treatment of severe atonic p.ostpartum hemorrhage and has few side effects. Its failure should raise suspicion that other pathologies such as lacerations and retained products of conception are responsible for the postpartum hemorrhage. This multicenter surveillance study was performed by site coordinators at 12 obstetrics units in collaboration with staff of the Medication Monitoring Program at The Upjohn Company. The 12 centers are Madigan Army Medical Center, Tacoma, Wash.; United Hospital, Grand Forks, N.D.; William Beaumont Army Medical Center, El Paso, Tex.; Dallas-Fort Worth Medical Center, Grand Prairie, Tex.; University of New Mexico, Albuquerque, N.M.; Tarrant County Hospital District, Fort Worth, Tex.; K. I. Sawyer Air Force Base Hospital, K. I. Sawyer Air Force Base, Mich.; Virginia Regional Medical Center, Virginia, Minn.; Beloit Memorial Hospital, Beloit, Wise.; Denver General Hospital, Denver, Colo.; St. Luke Hospital, Denver, Colo.; and Maricopa Medical Center, Phoenix, Ariz. REFERENCES 1. Bruce SL, Paul RH, Van DorstenJP. Control of postpartum uterine atony by intramyometrial prostaglandin. Obstet Gynecol 1982;59:47S-50S. 2. Clark SL. Phelan JP. Surgical control of OB hemorrhage. Contemp Ob/Gyn 1984;22:70-82. 3. Toppozada M, EI-Bossaty M, El-Rahman HA, El-Din AHS. Control of intractable atonic postpartum hemorrhage by IS-methyl prostaglandin F2a • Obstet GynecoI1981:58:32730. 4. Hayashi RH, Castillo MS, Noah ML. Management of severe postpartum hemorrhage due to uterine atony using an analogue of prostaglandin F 2a • Obstet Gynecol 1981;58: 426-9. 5. Hayashi RH, Castillo MS, Noah ML. Management of severe postpartum hemorrhage with a prostaglandin F2a analogue. Obstet Gynecol 1984;63:806-8. 6. Buttino L Jr, Garite TJ. The use of IS methyl F2a prostaglandin (Prostin ISM) for the control of postpartum hemorrhage. Am J Perinatol 1986:3:241-3. 7. Magil B. PGF 2a for postpartum hemorrhage-how well does it work? Contemp Ob/Gyn 1984:22:111-9. 8. Clark SL, Yeh S, Phelan JP, Bruce S. Paul RH. Emergency hysterectomy for obstetric hemorrhage. Obstet Gynecol 1984;64:376-80. 9. Lucas WE. Postpartum hemorrhage. Clin Obstet Gynecol 1980;23:637-46.