Search tips
Search criteria 


Logo of bloodtransLink to Publisher's site
Blood Transfus. 2010 April; 8(2): 132–136.
PMCID: PMC2851220

Press review


Alaia V, Frey BM, Siderow A, et al

A pair of naturally occurring antibodies may dampen complement-dependent phagocytosis of red cells with a positive antiglobulin test in healthy blood donors.

Vox Sang 2009; 97:338–47.

This is a fine study from a group led by Professor Hans U. Lutz, a well known expert on red cell senescence and the role of naturally occurring antibodies. Approximately 1 out of 10,000 blood donors has a positive direct antiglobulin test (DAT+) in tube. Apart from the occasional donor with signs of haemolysis, most DAT+ donors appear to be perfectly normal. The reason why DAT+ cells are not removed from circulation in these subjects is not understood.

Previous studies from other groups had shown that DAT+ cells from normal donors carried two antibody populations: an IgG autoantibody that precipitated band-3, and an anti-idiotype IgG against the former antibody. This resembles what happens on senescent red cells, as elucidated by Lutz’s group: oxidative damage or the binding of hemichromes causes the formation of band-3 oligomers; naturally occurring IgG anti-band-3 antibodies (band-3 NAb) are present at low concentrations in all normal subjects, but they do not bind significantly to native band-3 because of their low affinity; they do, however, react with band-3 oligomers because, in this case, they bind with both antigen-combining sites; the number of band-3 NAb per cell is not sufficient to enhance phagocytosis, but band-3 NAb are peculiar antibodies that form C3b2-IgG complexes which are potent precursors of the amplifying C3 convertase; this causes the deposition of large amounts of C3 and the subsequent removal by macrophages.

Since these phenomena occur on older red cells, the authors wondered whether IgG are located on senescent cells also in DAT+ donors. They separated red cells from normal and DAT+ donors by density, a proxy for age, and confirmed that a satisfactory separation had taken place by checking another age-related parameter, the band 4.1a/4.1b ratio. They found that DAT+ donors had a higher percentage of denser cells. Moreover, those cells had a higher band 4.1a/4.1b ratio (they were older) and more IgG molecules than the control cells. There were no qualitative differences between the two groups in the eluates of the age-separated fractions. Denser cells from both groups had a greater amount of IgG and this bound to band-3, IgG, F(ab’)2, and C3. Compared with control plasma, the plasma of most DAT+ donors contained more anti-F(ab’)2 and anti-C3. The denser cells of DAT+ donors were more efficiently phagocytosed than the corresponding fractions from normal donors but the opposite occurred in the presence of moderate amounts (12 mg/mL) of non-specific IgG in the liquid phase. The latter conditions more accurately resemble the in vivo situation.

Putting the pieces of the puzzle together, the authors concluded that the complement-dependent phagocytosis of older red cells from DAT+ donors was inhibited by the presence of anti-idiotypic and anti-C3 antibodies. The former probably act by impeding the binding of C3b to the band-3 NAb; the latter, by blocking the interaction of red cell-bound C3 with complement receptors on macrophages.

All natural antibodies cited in this study (anti-band-3, anti-F(ab’)2, and anti-C3) are normally found in all individuals. In DAT+ donors, they are only increased. It is tempting, therefore, to conclude that DAT+ donors are simply the extreme tail of a normal distribution. However, a recent, unrelated study (Rottenberg Y, Yahalom V, Shinar E et al. Blood donors with positive direct antiglobulin tests are at increased risk for cancer. Transfusion 2009; 49:838–42) suggests that DAT+ donors frequently develop cancer, particularly haematological malignancies (RR=8.3; 95%CI, 1.5–43.2). In the present study, 3/12 DAT+ donors were diagnosed with a disease, including a plasmacytoma, during a follow up of 1–11 years.

At present, DAT+ donors pose delicate problems of counselling, which cannot be solved until we know much more about their propensity to develop diseases. It should not, however, be difficult to acquire more data on this issue. Such data should also be invaluable in guiding further diagnostic work-up.

Yazer MH, Triulzi DJ

The role of the elution in antibody investigations.

Transfusion 2009; 49:2395–9.

In an era of cost-containment, even a time-honoured procedure, such as elution, cannot escape a critical reappraisal.

The authors decided to evaluate the contribution of elution to diagnostic work-ups. They collected information on elutions performed during two periods of almost one year in total. Neonates and recipients of Rh immune globulin were excluded. The indications for performing an elution were a newly diagnosed direct antiglobulin test (DAT) positive for IgG, an increase in strength of the DAT result in a recently (30 days) transfused patient, or a specific request by a clinician. The DAT was performed in a tube, with a polyspecific antiglobulin reagent followed, if necessary, by monospecific anti-IgG and anti-C3. Eluates were prepared by acid elution. Antibody screening was performed with low ionic or PEG-enhanced tube tests, in both plasma and eluate.

Of a total of 648 eluates, 244 (38%) were negative. As expected, the frequency of reactive eluates correlated with the strength of the DAT. Eluates were positive in 42% of microscopically positive DAT, 55% of weakly positive DAT, and 86% of 1–4+ DAT. Of the 404 reactive eluates, 284 contained warm autoantibodies, 30 anti-A or anti-B, 80 alloantibodies, and 10 a mixture of auto- and alloantibodies.

The eluates were defined as informative when they contained antibodies not detected in plasma. There were 82 such cases (13%). In 50 eluates, the newly found antibodies were panagglutinins. Only six of them were eventually detected in plasma. The other 32 cases included the 30 anti-A or anti-B, one anti-D, and one anti-E. The example of anti-D was detected in plasma 29 days later. The patient with anti-E was lost to follow-up. The last two patients had only a weakly reactive DAT (microscopically positive).

Of the 90 eluates containing alloantibodies, only two were associated with reported cases of delayed haemolytic reactions. In both cases, the eluate did not contain antibodies not found in plasma.

The authors recommend not performing elution when antibody screening tests in plasma are positive or when the DAT is only microscopically positive, unless immune haemolysis is strongly suspected. However, they admit that following this advice, they would have missed the only two alloantibodies, besides anti-A and anti-B, not already detected in plasma. Any conclusion drawn from this study is weakened by the lack of information about the clinical significance of the antibodies detected. Personally, however, I would concur with the authors.


Baumgartner JM, Silliman CC, Moore EE, et al

Stored red blood cell transfusion induces regulatory T cells.

J Am Coll Surg 2009; 208:110–9.

The majority of studies on post-transfusion immunomodulation have focused on the cellular portion of blood components, particularly the leucocytes, but many studies have also investigated soluble mediators. The authors tested the primary hypothesis that the supernatant of red cell concentrates induces regulatory T cells (Tregs) and the secondary hypotheses that storage accentuates this phenomenon and leucoreduction attenuates it.

T regs were defined as CD4+Foxp3+ cells. Mononuclear cells obtained from the peripheral blood of healthy volunteers were cultured for 5 days in the presence (20% v/v) of supernatant from red cell concentrates. The 20% concentration was calculated to correspond to a transfusion of ten red cell concentrates. Tregs were induced (5.6–6.2% versus 2.7%, p<0.05) in the presence, but not in the absence, of a secondary stimulus (anti-CD3). This was a specific up-regulation, because the supernatant had no effect on the non-specific T-cell activation caused by anti-CD3. The Treg induction was not influenced by the duration of storage (1 versus 42 days) or by leucoreduction. Those Tregs were immunosuppressive and inhibited (8–11% versus 39%) the proliferation of T-responder cells (CD4+CD25). The red cell storage solution alone was ineffective. Washing the red cells before the collection of the supernatant abolished the effect. The authors also measured a battery of cytokines, including interleukin-1β, −2, −4, and −10, interferon-γ, tumour necrosis factor-α, and transforming growth factor-β1, but the results suggested that these cytokines played no role in the induction of Tregs.

Tregs are potent immunosuppressive cells but their involvement in post-transfusion immunomodulation is purely conjectural. Moreover, the relevance of this in vitro study to the in vivo situation is not clear. The title of the paper makes an undue reference to a clinical setting ("transfusion") and erroneously mentions "stored", whilst the results negate any effect of storage.

Dineen SP, Roland CL, Toombs JE, et al

The acellular fraction of stored platelets promotes tumour cell invasion.

Surg Res 2009; 153:132–7.

Allogeneic transfusion is suspected of increasing tumour recurrence, although the available evidence is at best equivocal. Proposed mechanisms centre on immunosuppression. This paper suggests a different one.

Platelets contain a host of growth factors, which are released by α-granules upon activation. The authors studied the effect of the supernatant of stored platelet concentrates on pancreatic and breast cancer cell lines in vitro and on the growth of a pancreatic tumour in vivo in mice. The supernatant was obtained from apheresis platelets, leucoreduced on collection and stored for 7 days. As a control, the authors used the supernatant of the same platelet units after washing. The "unwashed" supernatant promoted the invasion of tumour cells (a human and a murine pancreatic cell line and a human breast line) in a modified Boyden chamber technique. The unwashed supernatant contained vascular-endothelial growth factor (VEGF). The addition of bevacizumab (a humanised monoclonal antibody against VEGF) to the supernatant reduced the invasion of the human pancreatic cell line but not of the murine one. The authors commented that VEGF is involved but is not the sole migration factor. In the in vivo study, a murine pancreatic tumour cell line was injected into the tail of the pancreas of mice. After 2 weeks, the animals were given an intravenous injection of supernatant (washed or unwashed) and after another 2 weeks they were sacrificed. The tumour size was not significantly different between the groups, but the microvessel density was increased in the group that received unwashed supernatant, indicating greater tumour angiogenesis in this group.

The authors commented that platelets in red cell concentrates probably degranulate during storage, releasing their content into the supernatant. I would add that to a certain extent, degranulation is also likely to occur during leucoreduction by filtration. If this were to be clinically important, lifelong transfusion-dependent patients, such as those with thalassaemia major, would be at a high risk of developing a cancer. Such an association was suspected in the past but not confirmed. This paper proposes a pathogenic mechanism, but does the disease exist?

Massive transfusion

Zink KA, Sambasivan CN, Holcomb JB, et al

A high ratio of plasma and platelets to packed red blood cells in the first 6 hours of massive transfusion improves outcomes in a large multicenter study.

Am J Surg 2009; 197:565–70.

This large multicentre retrospective study is a contribution to the recent controversy on the optimal plasma to red blood cell (P:R) ratio in massive transfusion.

The data come from 16 American level I trauma centres (level I refers to the American College of Surgeons classification and denotes the centres providing the highest level of care). Inclusion criteria were massively transfused trauma patients (ten or more units of packed red cells in the first 24 hours after injury), excepting those transferred from other hospitals, prisoners, children under 16 years old, pregnant women, burn-injured patients, patients who had 5 or more minutes of cardiopulmonary resuscitation before admission, patients undergoing an emergency room thoracotomy for blunt injury, and patients who died within 30 minutes of arriving in the centre. Data collected included the number and type of blood products transfused and the timing of transfusion (0–6 hours or 6–24 hours from admission), mechanism of injury, age, sex, injury severity score, initial Glasgow coma score, initial vital signs, initial laboratory tests, mortality by time from emergency room admission (0–6 hours, 6–24 hours, and >24 hours), and days on ventilation. The patients were divided into three groups, according to the P:R ratio and the platelet-to-red-cell (PLT:R) ratio during the first 6 hours after admission: low (<1:4), medium (1:4–1:1), high (>1:1). The primary outcome was inhospital mortality. Secondary outcomes were mortality in the first 6 hours, the number of red cell concentrates transfused in the first 24 hours, and ventilator-free days. The last outcome was included because of the concern that larger amounts of plasma or platelets can increase the frequency of transfusion-related acute lung injury (TRALI).

Overall, 466 patients were included in the study. Most of them were in the medium group as regards the P:R ratio and the low group as regards the PLT:R ratio. The groups did not differ significantly with regards to age, sex, vital signs, mechanism of injury, initial blood pressure or laboratory results (haemoglobin, platelet count, partial thromboplastin time, fibrinogen) but the median International Normalised Ratio was 1.5 in the group with high P:R ratio, versus 1.3 in the other groups (p=0.03). The rate of in-hospital mortality was 54.9%, 41.1%, and 25.5% in the groups with low, medium and high P:R ratios, respectively (p<0.04). The 6-hour mortality rates were 37.3%, 15.2%, and 2.0% in the same groups (p<0.001). Corresponding values for the three groups defined by PLT:R ratio were: in-hospital mortality: 43.7%, 46.8%, and 27.4% (p<0.03); 6-hour mortality: 22.8%, 19.0%, and 3.2% (p<0.002).

Patients in the groups with low and medium P:R ratios received a median of 18 red cell concentrates in the first 24 hours, while those in the high group received significantly less (13 units; p<0.001). The results in the groups defined by the PLT:R ratio were almost identical: 17.5 versus 13 units (p=0.008). The ventilator-free days were only calculated for the patients who survived to 30 days, to avoid the confounding factor of the early deaths. The ventilator-free days did not differ between the P:R ratio groups but were significantly fewer (p<0.004) in the group with a low PLT:R ratio (6 versus 9.9 and 9.1 days, respectively).

The optimal P:R ratio is one of the most controversial issues in massive transfusion in recent years. Many studies have documented an important survival advantage for patients who received approximately the same number of fresh frozen plasma and packed red cells (1:1 ratio). This has led many trauma centres to abandon the customary proportion of one unit of fresh frozen plasma every three units of red cell concentrates. However, no randomised clinical study has been published so far and the interpretation of the published ones is complicated by several methodological flaws. The most conspicuous is that the P:R ratio is calculated in the first 6–12 hours after injury but most haemorrhagic deaths in trauma patients occur during the first 2–3 hours. The first blood components transfused are always the red cell concentrates and therefore the early deaths are necessarily associated with a low P:R ratio.

The authors of the present study attempted to avoid this bias by excluding patients who died within 30 minutes of arrival (actually, only one patient was excluded) and calculating the ratio of the blood components transfused during the first 6 hours only. They also commented that there were no significant differences in the severity scores between the groups. However, one is led to wonder why the clinicians in charge administered more red cells in some patients and more plasma in others. They were not participating in a randomized clinical trial. A plausible explanation is that some patients were predominantly hypoxic, while others were not. In the latter patients, the clinicians directed their efforts to treating the coagulopathy because oxygen delivery was not compromised or had already been corrected. Therefore, a higher P:R ratio could just be a marker of a less severe clinical condition. In fact, this could also be an explanation for the lower consumption of red cells in this group.

A definitive answer will only come from prospective studies, but these will not be easy to set up.

Riskin DJ, Tsai TC, Riskin L, et al

Massive transfusion protocols: the role of aggressive resuscitation versus product ratio in mortality reduction.

J Am Coll Surg 2009; 209:198–205.

This study comes from a single level I trauma centre and underlines another potential confounding factor in the interpretation of the outcomes of massive transfusion protocols.

The authors compared the 2 years before and the 2 years after the implementation of a massive transfusion protocol (MTP). Salient characteristics of the MTP were a target ratio of plasma to red cell concentrates (P:R ratio) of 1:1.5 and a shorter time to have components available for transfusion. Upon activation of the MTP, blood components were sent to the patient’s bedside in sets composed of six units of red cell concentrates, four units of fresh frozen plasma, and one apheresis platelet concentrate. Activation of the MTP was recommended when more than four units of red cell concentrates (PRBC) were transfused in the first hour or when the expected transfusion requirements exceeded ten units in a 12-hour period. The goal of the resuscitation was haemodynamic stability. Haematocrit-based transfusion triggers were not used.

Inclusion criteria for the study were direct admission through the emergency department and the requirement of more than ten PRBC units over the first 24 hours. Exclusion criteria were preadmission care at an outside hospital, age less than 16 years, more than 5 minutes of pre-hospital cardiopulmonary resuscitation, emergency department thoracotomy performed for blunt injury, and pregnancy. The primary end-point was mortality (in-hospital, presumably; it was never defined in the paper).

In the Pre-MTP and Post-MTP periods, 40 and 37 patients, respectively, met the inclusion criteria. The groups were not significantly different with regards to age, sex, percentage of blunt trauma, various injury severity scores or Glasgow coma score. Surprisingly, the P:R ratio was 1:1.8 in both periods. However, the platelet to PRBC ratio was lower in the Pre-MTP period (1:1.7 versus 1:1.3, p=0.05). The authors commented that in their Institution dilutional coagulopathy was already aggressively managed before the implementation of the MTP. The numbers of the three blood components per patient were not different between the groups. There were significantly fewer deaths in the second period: 18 patients (45%) in the Pre-MTP and 7 (19%) in the Post-MTP (p=0.02). The time (minutes) to the first transfusion of cross-matched and/or type-specific blood component was significantly shorter in the second period: PRBC: Pre-MTP: 115, Post-MTP: 71 (p=0.02); plasma: 254 versus 169 (p=0.04); platelets: 418 versus 241 (p=0.01).

The authors performed a logistic regression analysis to identify variables associated with mortality. The only significant predictors of mortality were the number of PRBC (RR=1.21; 95% CI 1.06–1.38) and the number of platelet concentrates (RR=0.43; 95% CI 0.20–0.93). Therefore, PRBC were associated with a higher risk and platelets with a lower risk.

The authors attributed the significant decrease in mortality to the faster availability of blood components. From the paper, it may be inferred that the clinicians had only four PRBC available before the delivery of cross-matched/type-specific blood components, although this required almost 2 hours, on average, for the PRBC in the Pre-MTP period, and much more for plasma and platelets.

The authors also noted that previous studies were unable to separate the effects of changes in blood component ratio and time to transfusion. It is an interesting observation, that leads one to suspect that the improved survival described in previous studies, and attributed to the P:R ratio, could instead be due to the implementation of an MTP.

In conclusion, this study suffers from many limitations but has the unquestionable merit of drawing the attention to the advantages of a better collaboration between the emergency department and the transfusion service. In the really urgent cases, it is literally vital that access to blood components is prompt. The implementation of an MTP seems a sound way of communicating the degree of urgency, so that an appropriate response may be obtained.

Articles from Blood Transfusion are provided here courtesy of SIMTI Servizi