blood loss and the subsequent morbidity and
transfusion rates were similar, non-inferior, with
a trend to be lower after the switch to fibrinogen
concentrates. No adverse events were attributed
to fibrinogen supplementation. 23 Kikuchi et al
observed the successful correction of dilutional
coagulopathy in 12 patients undergoing a PPH
receiving fibrinogen concentrates. 24 Makino’s
series was larger: the Japanese five-year national
survey assessed 101 severe PPH patients who
received 3g fibrinogen concentrate (repeated once
in 17 patients). The initial plasma fibrinogen level
was 0.7g/l. Final blood loss was 4562 ± 3198ml.
The single dose infusion increased the plasma
fibrinogen level to 1.87 ± 0.72g/l; the
supplementation was not able to reach the target
of 1.5g/l for the two patients who died. 25 In this
PPH cases series, the dose of fibrinogen could
have been too low to correct the plasma level;
the timing of administration could have been
too late; the transfusion and coagulation data
collected were limited to red blood cells and
plasma while no information was described on
tranexamic acid and platelets.25 Mallaiah et al
compared two periods: a massive transfusion
protocol blind strategy during the first period
versus the use of an adapted dose of fibrinogen
concentrate targeted by point of care diagnosis
of hypofibrinogenemia during the second period.
They demonstrated a significant reduction in the
number of patients requiring more than 6 RBC
and in overall blood product requirements, in
maternal morbidity but not on hysterectomy
rate. 26
The first prospective, double-blind, placebo-
controlled, randomised study available – the
FIB-PPH trial27 – failed to show a difference
between placebo and fibrinogen in terms of
transfusion rate (21% compared with 22% in the
placebo group) in normo-fibrinogenaemic
patients (pregnancy and postpartum reference
values: 4.5–5.5g). In this trial, patients were
recruited in case of blood loss due to manual
removal of placenta (≥500ml), manual exploration
of the uterus after placental delivery (≥1000ml),
or C-section (perioperative blood loss ≥1000ml).
Patients were randomised to receive a single
uniform dose of 2g fibrinogen or placebo
systematically, that is, independently to the
laboratory- or the point of care plasma fibrinogen
level. There were limitations to the FIB-PPH trial:
the patients enrolled were non-severe and
normofibrinogenemic; the delay to administer
the product was long (more than one hour after
the diagnosis) and the population studied was
mostly surgical bleeding (C-section and genital
tract injury) as opposed to obstetrical bleeding
that leads to coagulation defects. Bleeding
probably stopped in a majority of patients with
a limited need for transfusion and no need for
invasive procedures in both groups. The selection
of a non-severe population with a low rate of
transfusion was mostly due to the difficulty in
recruiting and obtaining informed consents
during an ongoing bleeding; the primary
endpoint took into account only the transfusion
rate but not anaemia or haemoglobin drop in
non-transfused patients. Despite the pioneering
nature of the FIB-PPH trial, the scientific
questions on efficacy, safety, timing and dose of
fibrinogen concentrate in ongoing severe PPH
17
HHE 2018 | hospitalhealthcare.com
with hypofibrinogenemia have not yet been
answered. 27
Collins et al have published a randomised trial
hypothesising that an early fibrinogen
replacement, guided by thomboelastometric
test, reduces transfusion need and bleeding in
severe PPH. 28,29 They enrolled 55 women with
hypofibrinogenemia attested by a <12mm
FIBTEM A5 amplitude out of the 663 severe PPHs
recruited. The 55 enrolled patients were blinded
to fibrinogen or placebo. The median (25th–75th
centile) measured blood loss was 1450ml (1200–
1800ml) at enrolment and 1950ml (1500–2285ml)
when the studied medication was infused. 29
The adjusted incidence rate ratio (IRR) (95% CI)
for the number of allogeneic units transfused in
the fibrinogen group compared with placebo was
0.72 (0.3–1.7; p=0.45). In pre-specified subgroup
analyses, subjects who had a FIBTEM A5 <12mm
at the time of randomisation and who received
fibrinogen concentrate received a median of 1
(0–4.5) PRBC and had an additional 300ml
(100–350ml) blood loss whereas those who
received placebo also received 3 (0–6) units
of allogeneic blood products and had 700ml
(200–1550ml) additional blood loss (NS). There
was one thrombotic event in each group. 29
The authors concluded that infusion of
fibrinogen concentrate triggered by FIBTEM
A5 <15mm did not improve outcomes in PPH.
Pre-specified subgroup analyses suggested that
fibrinogen replacement is not required if the
FIBTEM A5 is>12mm or Clauss fibrinogen >2g/l,
but an effect below these levels cannot be
excluded.
However, in this trial, as in the data
of Wikkelso et al, 27 the recruitment of the
coagulopathic patients was difficult and women
were excluded if they declined transfusion, had
prenatally diagnosed placenta accreta, had
undergone an invasive procedure to control
bleeding, or if there was a clinical suspicion
of amniotic fluid embolism. Moreover, despite
a clear protocol design, the use of FFP was
encouraged in both groups and tranexamic
acid was given at randomisation. This may be
responsible for a confounding effect on bleeding
reduction. 20,21
The ongoing double-blind, placebo-controlled
RCT – the FIDEL trial – aims to assess the early
administration of 3g fibrinogen concentrates
versus placebo in more severe and ongoing PPH
resistant to the first-line uterotonics. 30 The
recruitment has been prolonged to achieve
inclusion of 462 patients and a 90% beta risk.
FIDEL is an efficacy trial. The primary efficacy
variable is a binary composite endpoint (failure
versus success). Failure is defined when a patient
loses at least 4g/dl Hb, and/or requires the
transfusion of at least two units of packed RBCs
within 48 hours following the administration of
investigational medicinal product. The Hb
reference level is defined as the last Hb value
recorded within the third trimester of pregnancy.
To avoid or minimise bias and/or centre effect, the
primary endpoint has been defined independently
of therapeutic obstetrical interventions and/or
clinical practices such as surgical ligation or
embolisation. All centres currently use a French
guidelines’ algorithm for PPH management. 31 The
secondary objectives of this trial are to evaluate