Abstract
Abstract
Cell salvage is widely used in many surgical specialties to reduce the use of allogeneic blood, which is associated with well-recognized significant complications such as incorrect transfusion, acute and delayed immunological transfusion reactions and transfusion-transmitted infection. Theoretical concerns over the use of cell salvage in obstetrics have focused on the risk of Rhesus immunization and the risk of contamination of the cell-saved blood with traces of amniotic fluid. This article examines the in vitro and in vivo evidence for the safety of the technique, presents current usage figures from Wales and the entire UK, and discusses the current perception of amniotic fluid embolus as an anaphylactoid reaction. It suggests that patients will have to choose between the risks of allogeneic or cell-saved blood, and concludes that the balance of evidence is now in favour of cell salvage in obstetrics.
Introduction
Cell salvage is a well-established technique, providing a method of perioperative transfusion with autologous blood (the patient's own, shed blood) instead of homologous transfusion with allogeneic donor blood, thereby avoiding many of the risks of blood transfusion. Its use specifically in obstetrics has been controversial because of two theoretical risks.
First, there is a fear of incomplete cleaning of the blood, leading to possible contamination of the re-infused blood with amniotic fluid. The classical view that exposure of maternal blood to amniotic fluid will inevitably cause the catastrophe of ‘amniotic fluid embolus’ (AFE) had discouraged the development of cell salvage in obstetrics for many years. There are now good reasons to re-evaluate the nature of AFE.
Secondly, cell salvage in obstetrics presents a theoretical risk of Rhesus immunization of an Rh negative mother with red cells from an Rh positive fetus.
These two theoretical risks must be balanced against the risks of allogeneic transfusion. This article examines the current state of knowledge in this field and attempts to weigh these relative risks.
What are the risks associated with allogeneic blood transfusion?
The hazards of allogeneic blood transfusion have been quantified in UK since 1996, when Serious Hazards of Transfusion (SHOT) was established. The first report from this voluntary, anonymous reporting system analysed serious incidents in the five-year period, 1996–2001. 1 This showed that of 1148 serious events reported, there were 699 incorrect blood transfusions given, 417 adverse immunological reactions including 70 transfusion-related acute lung injury (TRALI) and 32 transfusion-transmitted infections (TTI).
Getting the wrong blood
Of the 699 incorrect transfusions, 161 were ABO compatibility errors which resulted in 11 deaths and 60 cases of major morbidity.
In addition, there were 73 Rhesus errors which involved 17 women of childbearing age.
Of the 17 million blood components issued this translated to:
Risk of receiving wrong component = 1 in 25,000 Risk of ABO incompatibility (161) = 1 in 106,000 Risk of major complication (166) = 1 in 103,000 Risk of death (41) = 1 in 415,000 NHS Trusts where blood is transfused should ensure that patients are aware of the option of autologous transfusion and consider the introduction of perioperative cell salvage systems.
Although the absolute risk to any individual recipient is low, the finding of this level of error in the administration of blood transfusion was a shock to most clinicians. In 2002, the Department of Health published the document Better Blood Transfusion
2
which contains lengthy recommendations to make allogeneic transfusion safer and includes the advice that:
Unfortunately, the most recent SHOT Report 20051 indicates 485 ‘incorrect transfusions’ with two deaths — in one year, an increase of 246% since 1996–2001 — although ABO incompatibility errors are much reduced.
Cell salvage removes the risk of receiving the wrong blood, as the re-infused blood is the patient's own, given at the bedside, during or immediately after the loss.
Transfusion-transmitted infections
The 1996–2001 SHOT Report 1 also studied TTI. There were 39 proven transmitted infections, which included eight Hepatitis B and two Hepatitis C.
The existence of the Hepatitis C virus (HCV) was not established until 1989, and a reliable screening test has only been available in UK since 1991. 3 After 1991, HCV positive donors were identified and excluded from donation, with the result that only two cases of transfusion-transmitted HCV infection have occurred since the SHOT reports began in 2001. 1 However, anyone receiving donor blood prior to this date may have been exposed to HCV, and current estimates suggest that there are 200,000 people in the UK with the infection, of which 20% will progress to cirrhosis or hepatocellular carcinoma. Although most of these are related to intravenous drug abuse, an unknown number are directly due to blood transfusion. The UK is the only developed country where the rate of chronic liver disease is increasing. 3
Prion transmission via blood transfusion was not suspected until 2003, and it has since become clear that there have been four cases of probable variant Creutzfeldt-Jakob disease (vCJD) infection via the blood supply. 3 The implication for the future of the blood supply and the risk to which the transfused population may have been exposed are not yet clear. Unlike HCV, there is still not a test or a treatment for vCJD, and a ‘prion filter’ is not yet available.
After the public health disaster of HCV and the current uncertainty about the ongoing degree of transmission of vCJD, it is reasonable to ask what other ‘as yet unknown’ viruses or prions may be in donor blood.
Cell salvage removes the risk of TTI, as the recipients receive only their own blood.
Immunological reactions
The 1996–2001 SHOT Report 1 also identified 410 immunological complications in the five-year period, including 70 cases of TRALI, of which 20 (28.5%) were fatal and a further 49 (70%) involved major morbidity. TRALI is a recently recognized serious immunological reaction to allogeneic blood.
The 2005 SHOT Report 1 identifies 121 immunological reactions including 23 TRALI cases with two deaths. For TRALI, this is an increase of 64% in the 1996–2001 figures.
The 2005 Report states ‘there were no immunological reactions with reinfusion of autologous blood’ (Table 1). 1
Cumulative mortality/morbidity from allogeneic blood transfusion 1996–2005 (serious hazards of transfusion)
IBCT, incorrect blood component transfused; ATR, acute transfusion reaction; DTR, delayed transfusion reaction; PTP, post-transfusion purpura; TA-GvHD, transfusion-associated graft versus host disease; TRALI, transfusion-related acute lung injury; TTI, transfusion–transmitted infections
*Major morbidity defined as one or more of the following: ITU admission/ventilation; dialysis and/or renal impairment; major haemorrhage from transfusion-related coagulopathy; intravascular haemolysis and potential risk of D sensitization in female of child-bearing age
†Excludes seven cases from 1998 to 1999 which were unclassified
How does this compare with the risks of cell salvage?
There are no proven deaths due to cell salvage. Cell salvage is often used in emergency surgery on the sickest cases (e.g. ruptured aortic aneurysm), which die because of their underlying condition, but there have been no cases in any surgical specialty where cell salvage has been implicated as the primary cause of death.
Cell salvage in obstetrics
In the UK, there are about 1400 cases of major haemorrhage in obstetrics per year,
4
and massive obstetric haemorrhage remains a leading cause of maternal death. In the 2000–2002 Report of the Confidential Enquiries into Maternal Deaths (CEMACH),
5
there were 17 deaths due to haemorrhage — this incidence is more than double that of the 1997–1999 triennium, when there were seven deaths from haemorrhage. This is a cause for great concern. Obstetric bleeding consumes about 3–5% of the UK blood supply,
6
which makes obstetric surgery particularly suitable for cell salvage. In obstetrics there are two special concerns:
Fetal red cell contamination with Rhesus immunization: The cell saver cannot distinguish fetal from maternal red cells, so any fetal red cells shed and aspirated into the cell saver will be re-transfused. In Rhesus negative mothers, this may increase the dose of Anti-D immunoglobulin required to prevent Rhesus immunization. Between 2 and 19 ml of fetal red blood cells may be re-transfused,
7
requiring 500–2500 IU Anti-D and Kleihauer testing should be performed as soon as practicable. However, it must be borne in mind that all Rh negative mothers who bleed antenatally or who undergo Caesarean section (CS) will require Anti-D, and the use of cell salvage may increase the dose required.
Amniotic fluid: is it removed by cell salvage? The key concern is whether there is a risk of re-transfusing amniotic fluid suctioned from the surgical site, or can we be confident that the cell saver removes it? What would actually happen if some residual amniotic fluid was transfused into the mother by this route — would she undergo ‘amniotic fluid embolism’?
Over the past 10 years, in vitro work has attempted to answer this question by studying the ability of the technical equipment to clear the various elements of amniotic fluid from blood. Meanwhile, the technique has been adopted widely in UK by enthusiastic clinicians who have accepted the growing body of clinical evidence of safety and ‘voted with their feet’!
For example, in Wales between April 2006 and June 2007, there were 110 obstetric cell salvage procedures carried out, and the blood was returned to the patient in 29 of these (WBS Cell Salvage Database, courtesy of Hannah Grainger).
A UK National Survey of Obstetric cell salvage use in the UK 2005–2006 conducted by the Obstetric Anaesthetists Association 8 revealed that 38% of UK maternity units were using obstetric cell salvage at that time, and that 12% include cell salvage in their unit's Major Obstetric Haemorrhage protocol. In this survey, barriers to the use of cell salvage were clearly identified as predominantly being lack of equipment and training — not concerns over safety.
In vitro clearance studies
The ability of cell salvage combined with modern leucocyte depletion filters (LDF) to remove amniotic fluid from blood aspirated from the surgical field at CS has been extensively studied.
A randomized controlled trial of cell salvage with a power of 80% to prove that the technique does not increase the incidence of AFE by up to five-fold would require up to 275,000 patients and is therefore unachievable.
However, every element of amniotic fluid that has been studied both in vitro and in vivo has been effectively removed by the combination of cell salvage and LDF. 7,9,10 Using the Haemonetics Cell Saver 5 in combination with the Pall RS leucocyte depletion filter (Leuco Guard RS, Pall Biomedical Products Co., East Hills, New York, USA), Waters achieved complete removal of fetal squames and phospholipid lamellar bodies from blood aspirated from the surgical field in patients at CS. 10
However, the problem with all clearance studies is the difficulty in ‘proving a negative’ — to prove that something is ‘not there’ requires a test with 100% sensitivity, and all current methods of detection of elements of amniotic fluid would be open to the criticism that the test simply is not good enough.
There are now published studies describing the use of cell salvage in over 400 clinical obstetric cases with no evidence of harm, 11–15 and two editorials concluding that the technique is safe and should be encouraged, 16,17 as well as a large and increasing body of individual ‘personal experience’ cases that are not published.
Prior to 2000, the LDFs were not available, so all obstetric patients cell saved before this will presumably have had some elements of amniotic fluid re-transfused.
There is one published letter that refers to a Jehovah's Witness with severe pre-eclampsia and HELLP syndrome (haemolysis, elevated liver enzymes, low platelets) who died following CS in which cell salvage was used. 18 Her preoperative Hb was 7.1 g/dl, platelet count 48 × 109 and she required emergency CS under general anaesthesia. She suffered hypoxic cardiac arrest a few minutes after endotracheal extubation. No further clinical or pathological details are given, and causes of death other than AFE were not considered. It is not possible to say whether cell salvage contributed to the death of this critically ill parturient, but most obstetric anaesthetists do not regard this as a death due to cell salvage.
Does amniotic fluid routinely enter the maternal circulation?
Given the successful obstetric cell salvage outcomes prior to the use of filters, we might ask: does amniotic fluid frequently — or even routinely — enter the maternal circulation and cause no ill effect?
Until relatively recently, the presence of fetal squames in the maternal blood was regarded as a marker of AFE, and the mechanism of death was assumed to be a particulate embolus of fetal squames. This was first questioned when the technique of pulmonary artery flotation catheterization in the 1980s enabled sampling direct from the pulmonary circulation of ‘well’ parturients with cardiac disease. Many of these patients had abundant ‘fetal squames’ in their central circulation, leading Clarke to suggest that this was a normal finding, 19 probably due to skin contamination during cannulation. Indeed, the Confidential Enquiries into Maternal Deaths in the UK has abandoned the histological finding of fetal squames in the lungs as a diagnostic criterion in AFE.
Fortunately, there is a more specific marker for amniotic fluid contamination. Waters et al. 10 showed that phospholipid lamellar bodies were present in 100% of CS patients at the time of placental separation. Lamellar bodies are fetal phospholipids from the developing fetal lung. These patients had femoral venous catheters sited preoperatively, with the catheter tip lying in the vena cava at the level of the uterine veins, and blood was sampled at the time of delivery. The implications of this paper for our understanding of AFE are profound — all 15 of these study patients had amniotic fluid in their blood, and were physiologically and clinically normal.
Current evidence supports the view that the term ‘AFE’ is a misnomer, and that the clinical syndrome of peripartum cardiovascular collapse, hypoxia and disseminated intravascular coagulation that occurs in 1 in 8000 to 80,00020 is a rare anaphylactoid reaction to unknown mediators.
Intuitively, it seems reasonable to remove amniotic fluid and placental tissue before aspirating to the cell saver, so as to minimize the volume of material to be removed by the filter. Conventional suction is used to aspirate all visible amniotic fluid after delivery of the baby before changing to cell salvage suction. However, the author is aware of many cases where cell salvage has been used throughout the procedure (e.g. immediate severe bleeding from placenta accreta in a Jehovah's Witness), and also of many cases where cell salvage has been used without LDF — both with no ill effect (personal communications).
Conclusions
Ultimately, patients will have to choose between the risks of allogeneic blood and the risks of the alternatives. Fears over the safety and sustainability of the blood supply and relative costs may influence this decision at Department of Health and Trust management level. At present the evidence is in favour of cell salvage in obstetrics, and this is reflected by its rapidly increasing use in the UK.