Alfirevic, Zarko MD, FRCOG; Elbourne, Diana PhD; Pavord, Sue MBChB, FRCP; Bolte, Antoinette MD; Van Geijn, Herman MD; Mercier, Frederic MD; Ahonen, Jouni MD; Bremme, Katarina MD; Bødker, Birgit MD; Magnúsdóttir, Ebba Margrét MD; Salvesen, Kjell MD; Prendiville, Walter MAO, FRCOG; Truesdale, Ann BSc; Clemens, Felicity MPhil; Piercy, Deborah BSc; Gyte, Gill BSc
Major primary postpartum hemorrhage is a life-threatening complication of labor often affecting healthy mothers without warning. The definition is usually based on an estimated blood loss of 1 L or more during the first 24 hours postpartum,1 with the incidence in Europe of approximately 0.5%.2 Treatment strategies focus on prompt hemostasis and cardiovascular support with intravascular transfusions of large quantities of blood products. In recent years, misoprostol,3 vessel embolization by interventional radiology,4 and uterine hemostatic sutures5 have been added to the management protocols. However, the morbidity and mortality from postpartum hemorrhage remains unacceptably high in both the developed and developing world.6,7
Whatever the primary cause of hemorrhage, it is important that hemostatic measures are not compromised by an inadequate clotting mechanism. For patients with hemophilia complicated by inhibiting antibodies, recombinant activated factor VII (rFVIIa, NovoSeven, Novo Nordisk, Bagsvaerd, Denmark) has been licensed for use during bleeding episodes and surgical procedures. Recombinant activated factor VII initiates coagulation when tissue factor (TF) has been exposed at the site of injury.8,9 Factor VIIa-TF complex mediates activation of factor X, which subsequently converts prothrombin into thrombin. The limited amount of thrombin generated through this pathway activates factors V, VIII, XI, and platelets, enhancing its own production in a powerful positive feedback loop. Recombinant FVIIa is capable of binding directly to the surface of activated platelets, further mediating activation of coagulation at the local site of injury.10,11 The prohemostatic effect and restriction to the local site has led to the increased use of rFVIIa in patients with massive hemorrhage who have preexisting normal coagulation.
There have been several hundred cases of rFVIIa administration to women with severe primary postpartum hemorrhage worldwide, most of which remain unreported. These numbers seem to be increasing exponentially with the tendency to use rFVIIa earlier and earlier in the management of primary postpartum hemorrhage. The use of rFVIIa is likely to rise further, following a continuing trend from opinion leaders to promote its off-license use.12 This is a cause for concern given the lack of evidence on its effectiveness and safety for this indication, and the cost of the drug (between $9,000 and $18,000) depending on the women's estimated weight and whether one or several doses are used).
The aim of the study reported here was to collect a prespecified set of data from nine European countries for all cases of obstetric hemorrhage between 2000 and 2004 in which rFVIIa was used. In the absence of prospective studies with adequate controls, we aimed to provide a methodologically robust description of practice in the early phase of rFVIIa for primary postpartum hemorrhage and a solid evidence base for design of epidemiologic studies and randomized trials in this area.
MATERIALS AND METHODS
The methods for obtaining the data varied slightly among the participating countries. In the United Kingdom, the Northern Europe factor 7a in Obstetric Hemorrhage Registry was approved by Metropolitan (London) Multicenter Ethics Committee in 2004, with the agreement that patient consent was not required if all data sent to the coordinating center were made anonymous at the local hospital. In 2005, a letter was sent to all U.K. maternity hospitals explaining the study and enclosing a Registry questionnaire, asking whether rFVIIa had been given for primary postpartum hemorrhage between 2000 and 2004 inclusive. This letter was sent to a named local obstetrician, but also copied to the pharmacist and the hematologist to avoid missing any relevant use of rFVIIa. The study was also publicized in relevant conferences and journals. Once a hospital was identified as having used rFVIIa for primary postpartum hemorrhage, the respondent was sent a Case Record Form to complete for each woman.
The Northern Europe factor 7a in Obstetric Hemorrhage Case Record Form asked for brief sociodemographic details and past medical and obstetric history, followed by details of the progress and management of labor in which the primary postpartum hemorrhage occurred. These included complications of labor and management, details and cause(s) of hemorrhagic complications and any accompanying morbidity, standard management applied, rFVIIa therapy (dosage used, timing of any response), maternal outcomes, and any adverse events after rFVIIa use. A number of reminders, including postal, e-mail, and telephone contacts were used to obtain both the Registry data and the Case Record Forms. A small payment was offered for completed Case Record Forms.
In Denmark, France, Iceland, Norway, and Sweden, the relevant national approvals were obtained for hospitals to be contacted as described above. The English version of the Case Record Form was used except in France and Denmark, where the national co-coordinators arranged for a translation. In Denmark, in two cases a full copy of the patient hospital records, including anesthetic records and laboratory tests were sent to the national co-coordinator, who filled the Case Record Forms based on this material. A national survey had already been conducted in Finland,13 and the national co-coordinator agreed to contribute the data using Northern Europe factor 7a in Obstetric Hemorrhage Case Record Forms. Similarly in the Netherlands, a survey had been undertaken up to July 2004. A research fellow visited each hospital where cases were identified to complete the Case Record Forms. In the Republic of Ireland, the national co-coordinator contacted all maternity units, and because no use of rFVIIa was reported, no further action was taken.
Administration of rFVIIa was defined as treatment when, after primary postpartum hemorrhage, standard treatment was considered to have failed, including administration aiming to prevent cesarean hysterectomy, or during cesarean hysterectomy on an unstable patient with continuous excessive bleeding. Administration was defined as secondary prophylaxis if rFVIIa was used to support other interventions that were considered successful (including administration during an uncomplicated cesarean hysterectomy). Clinicians were asked to describe subjectively the effect of rFVIIa administration using two mutually exclusive categories: 1) bleeding reduced or 2) bleeding unchanged or worse.
Data were entered and processed at the Medical Statistics Unit, London School of Hygiene and Tropical Medicine. We used a Mantel-Haenszel χ2 test to estimate odds ratios for all different causes of bleeding by type of treatment (treatment or secondary prophylaxis).
Table 1 shows the number of hospitals providing information regarding the number of hospitals using rFVIIa for primary postpartum hemorrhage and the number of cases in each participating country (Fig. 1). Sixty-five of the 531 hospitals with known usage status (12.2%) reported use of rFVIIa for primary postpartum hemorrhage on 128 women between 2000 and 2004 inclusive. The number of reported cases increased from three in 2001 and seven in 2002 to 31 in 2003 and 65 in 2004 (Fig. 1). Of 16 cases of secondary prophylaxis, thirteen (81%) were administered in 2004. Although we did not manage to make direct contact with 634 maternity units (54.4%), it is unlikely that nonresponding hospitals used rFVIIa in large numbers. This anecdotal evidence is based on the communications with the worldwide supplier of the drug and local informal contacts by the national coordinators. A total of 113 Case Record Forms have been returned (88%), of which 97 (86%) were classified as treatment, 16 (14%) as secondary prophylaxis. Five treatment cases were excluded from further consideration, four because of the secondary primary postpartum hemorrhage (uterine atony 8 days after vaginal delivery; continuous bleeding 9 days after hysterectomy and four subsequent operations; infection and disseminated intravascular coagulation (DIC) 14 days after cesarean delivery; suspected endometritis 16 days after cesarean delivery); and one where rFVIIa was given at the time of cesarean delivery to correct DIC caused by antenatal sepsis.
Table 2 describes the clinical status of the 108 women before rFVIIa administration. Most women (62%) had an uncomplicated pregnancy, with no relevant obstetric or medical history. Cesarean delivery was the most common mode of delivery (49% in the treatment and 50% in the secondary prophylaxis group), with uterine atony as the most common main cause of bleeding (60 of 108 cases (56%). There was no statistically significant difference in the main causes of bleeding between the treatment and secondary prophylaxis group.
Nearly all women (95%) had received at least one medical treatment for primary postpartum hemorrhage, mainly oxytocin, ergometrine, or prostaglandins before rFVIIa administration. Twenty-five women (22%) received misoprostol (mainly rectally), and the same number received antifibrinolytic therapy. A variety of hemostatic interventions were attempted for 95 women (87%). Twenty-one women had only one intervention (16 manual uterine explorations, three uterine packing, two embolizations) whereas the rest had multiple interventions, of whom 34 had a hysterectomy.
Women in the treatment group (n=92) had a median blood loss of 5.8 L before rFVIIa administration compared with 2.5 L in the secondary prophylaxis group. All but six of the 105 women for whom information was available received a transfusion of packed red cells (median 13 units in the treatment and 6 units in the secondary prophylaxis group). In addition, 95 women (86 in the treatment group, nine in the secondary prophylaxis group) received transfusion of other blood products in varying quantities (Table 2). Clinical disseminated intravascular coagulopathy was reported in 70% of cases in the treatment group and 25% in the secondary prophylaxis group.
Most women (88 of 108, 81%) received a single dose (Table 3). The commonest recorded dose was 7.2 mg or less (68 of 75, 91%) which is 90 mcg/kg or less for a woman weighing up to 80 kg. Clinicians noted improvements after a single dose for 80% of the women in the treatment group, and for 75% in the secondary prophylaxis group. Four more women responded to multiple doses, three in the treatment and one in the secondary prophylaxis group (Fig. 2). Recombinant activated factor VIIa was thought to have failed in 15 cases (13.8%).
The hemoglobin levels before and after the administration of rFVIIa are shown in Figure 3. Fifty-three paired samples where clinicians felt that bleeding was reduced showed increase in hemoglobin of at least 1 g/dL in 27 cases (51%), no significant change (less than 1g/dL) in 17 cases (32%), and drop in hemoglobin more than 1 g/dL in 9 cases (17%). In 6 of 12 cases hemoglobin had risen by more than 1g/dL despite the fact that clinicians felt that there was no clinical improvement after rFVIIa (Fig. 4).
Very few adverse events were noted as possibly related to rFVIIa administration. There were four cases of thromboembolism (two women developed pulmonary embolism within 1 week of birth, one had bilateral ovarian vein thrombosis 4 weeks after primary postpartum hemorrhage, and one woman developed a thrombus involving jugular and subclavian vein, right upper arm, and axilla, which was not thought to be related to the rFVIIa use). One woman was diagnosed with myocardial infarction, although she experienced cardiac arrest before rFVIIa was given. There was a suspected allergic reaction (skin rash) that prompted the decision to perform a hysterectomy rather then administer second dose of rFVIIa.
Although 23% of the women in this cohort had no further reported complications (Table 4), 75% were admitted to intensive care units, and five women died (all in the treatment group). Complications occurring before death included renal failure (n=1), respiratory failure (4), other organ failure (1), cardiac arrest (4), DIC (4), and sepsis (1). The causes of death were recorded as hemorrhage from rupture of splenic artery aneurysm, irreversible hypovolemic shock, hemorrhage from liver, Group A β hemolytic streptococcal sepsis, and postresuscitation neurologic damage.
This large case series describes the use of rFVIIa for the treatment and secondary prophylaxis of severe primary postpartum hemorrhage. Most of the reported use was in cases where all conventional measures were failing. The distinction between treatment and secondary prophylaxis in the context of major primary postpartum hemorrhage was made retrospectively by the clinicians involved in the patients' care using the definitions given in Materials and Methods above). Although we accept the arbitrary nature of this distinction, we believe that future studies should attempt to differentiate between administration when standard treatment has failed and “prophylactic” use when the conventional treatment is working, but the risk of further blood loss is high.
Clinical reports and hematologic data suggest very marked improvement for more than 80% of the women after rFVIIa administration and very few adverse effects. However, in the absence of an adequate control group, any attempts to comment on the true effectiveness of the drug should be resisted. Even with rFVIIa, 82 women were admitted to intensive care units. Five of the 113 women died, and a further 54 women (53%) needed laparotomy (including hysterectomy in 33 cases).
It is clear from our data that the use of rFVIIa has increased exponentially in Northern Europe over the 5-year period between 2000 and 2004. This is particularly true for prophylactic use of rFVIIa, with 13 of 16 cases recorded in 2004. Although formal data collection ended in 2004, there is anecdotal evidence from these countries that this upward trend is continuing, despite lack of objective information about effectiveness or cost-effectiveness for rFVIIa in primary postpartum hemorrhage. It is important to note that clinical impression of treatment success (or failure) is likely to be biased in any open label evaluation. However, our data suggest that changes in standard hematologic indices do not reflect accurately the rapidly changing clinical picture in these difficult situations where numerous medical and surgical interventions are employed simultaneously. Clearly, formal assessment of effectiveness needs to rely on hard clinical outcomes, including hysterectomy and serious maternal morbidity. Reduction in the amount of transfused blood products is important, but would have to be assessed in the context of detailed health economic evaluation.
The current situation poses significant challenges for clinicians worldwide. Anecdotal evidence suggests that the window of opportunity for clinical trials of rFVIIa in women who continue to bleed after postpartum hysterectomy may be disappearing as, increasingly, the drug is seen as an essential intervention when everything else has failed. The equipoise may have shifted to cases where rFVIIa is given to prevent hysterectomy rather than as a desperate measure when nothing else seems to work. However, the public is increasingly uneasy about these uncontrolled experiments without knowledge of the true advantages and disadvantages. Although the data on drug-related adverse effects from this series seem reassuring, the risk–benefit ratio may change if rFVIIa is given in mild to moderate cases of primary postpartum hemorrhage.
There is a continuing need for methodologically robust reporting of nationally and internationally collected observational data to monitor the changing situation.14 To answer the important question about whether rFVIIa is effective and cost-effective, randomized controlled trials are essential. However, the ethical issues related to consent and the use of placebo and the difficulty of obtaining commercial funding given the relative rarity of such severe primary postpartum hemorrhage, at least in developed countries, pose huge challenges. Nevertheless, a publicly funded trial is currently recruiting in France.15
1. Mousa HA, Alfirevic Z. Major postpartum haemorrhage: survey of maternity units in the United Kingdom. Acta Obstet Gynecol Scand 2002;81:727–30.
2. Zhang W, Alexander S, Bouvier-Colle, Macfarlane A, MOMS-B Group. Incidence of severe pre-eclampsia, postpartum haemorrhage and sepsis as a surrogate marker for severe maternal morbidity in a European population-based study: the MOMS-B survey. BJOG 2005;112:89–96.
3. Hofmeyr GJ, Walraven G, Gulmezoglu AM, Maholwana B, Alfirevic Z, Villar J. Misoprostol to treat postpartum haemorrhage: a systematic review. BJOG 2005;112:547–53.
4. Boulleret C, Chahid T, Gallot D, Mofid R, Tran Hai D, Ravel A, et al. Hypogastric arterial selective and superselective embolization for severe postpartum hemorrhage: a retrospective review of 36 cases. Cardiovasc Intervent Radiol 2004;27:344–8.
5. El-Hamamy E, B-Lynch C. A worldwide review of the uses of the uterine compression suture techniques as alternative to hysterectomy in the management of severe post-partum haemorrhage. J Obstet Gynaecol 2005;25:143–9.
6. Confidential Enquiry into Maternal and Child Health. Why mothers die? 2000-2002 The Sixth Report of the Confidential Enquiries into Maternal Deaths in the United Kingdom. London (UK): RCOG Press; 2004.
7. Khan KS, Wojdyla D, Say L, Gulmezoglu AM, Van Look PF. WHO analysis of causes of maternal death: a systematic review. Lancet 2006;367:1066–74.
8. Hedner U. NovoSeven as a universal haemostatic agent. Blood Coagul Fibrinolysis 2000;11:S107–S111.
9. Rapaport SI, Rao LV. The tissue factor pathway: how it has become a “prima ballerina.” Thromb Haemost 1995;74:7–17.
10. Hoffman M, Monroe DM, Oliver JA, Roberts HR. Factors IXa and Xa play distinct roles in tissue factor-dependent initiation of coagulation. Blood 1995;86:1794–180.
11. Weiss HJ, Lages B. Evidence for tissue factor-dependent activation of the classic extrinsic coagulation mechanism in blood obtained from bleeding time wounds. Blood 1988;71:629–35.
12. Branch DW, Rodgers GM. Recombinant activated factor VII: a new weapon in the fight against hemorrhage. Obstet Gynecol 2003;101:1155–6.
13. Ahonen J, Jokela R. Recombinant factor VIIa for life-threatening post-partum haemorrhage. Br J Anaesth 2005;94:592–5.
14. Sobieszczyk S, Breborowicz GH, Platicanov V, Tanchev S, Kessler CM. Recombinant factor VIIa in the management of postpartum bleeds: an audit of clinical use. Acta Obstet Gynecol Scand 2006;85:1239–47.