Passive IgE-sensitization by blood transfusion

Abstract

Background:  To study the mechanisms of passive sensitization of patients receiving plasma containing IgE antibodies to a defined allergen.

Methods:  When required for medical reasons, regular donor plasma with IgE antibodies to timothy grass allergen (8–205 kU_A/l), was given. Kinetics of IgE antibodies in the recipients’ serum and his/her basophil allergen threshold sensitivity, CD-sens, was monitored up to 2–3 weeks after transfusion. The IgE antibodies were quantitated by ImmunoCAP. The CD-sens in plasma recipients, determined by CD63 up-regulation, was measured by flow cytometry and compared to CD-sens of patients with allergic asthma and/or rhinitis.

Results:  There was a significant correlation (r = 0.98; P < 0.001) between amount of IgE antibody given and recipient serum peak concentration. The T_1/2 for IgE antibody in circulation was 1.13 days (95% confidence limit 0.35–1.91 days). The recipients became CD-sens positive already 3 h after transfusion. The CD-sens peak was observed after 3.4 days and the value were correlated (r = 0.68; P < 0.02) to the amount of IgE antibody transfused and were of the same magnitude as found in allergic patients. The T_1/2 of CD-sens indicated two populations of basophils; one with a CD-sens decrease T_1/2 of 4 days and one of 10 days.

Conclusion:  Transfused IgE antibodies will sensitize mast cells and basophils to CD-sens levels similar to those of allergic patients. The recipients expressed ‘slow’ or ‘rapid’ CD-sens decline, indicating two different basophil populations. After transfusion of plasma with >10 kU_A/l IgE antibody the recipient could have allergen reactive basophils for up to 7 weeks.

In 1919 Ramirez reported (1) the first case of allergy passively transferred by a blood transfusion. The patient was suffering from aplastic anemia and received blood regularly, and usually without any reactions. However, this particular time, when riding a horse carriage in Central Park a few days after the transfusion, the patient developed asthma. Dr Ramirez noticed the significance of this special event and when tracing the blood donor found that he was suffering from asthma due to horse allergy. Since this first description of transfer of an allergy, a few reports have followed. In 1941 Loveless reported skin and mucosal sensitivity to ragweed after transfer of blood from ragweed allergic patients shown to have circulating ‘reagins’ to ragweed (2). The nonallergic recipients became skin test positive for quite long times, in two of the five cases for 7 and 20 weeks, respectively. A couple of other cases have also been reported (3, 4).

The blood transfusion specialty has, through questionnaires and by simple case histories taken by blood center personnel, tried to identify and eliminate allergic, potentially dangerous blood donors. However, this procedure has not been very successful; we found that 23% of the donors were IgE-sensitized and had significant levels of IgE antibodies to common allergens (5). Thus, the risk of passive sensitization by blood and plasma transfusion is a reality. However, it also opens an opportunity to study serological and inflammatory aspects of human IgE-sensitization in vivo.

In the 1960s, before the discovery of the central role of IgE (6) in allergic inflammation, extensive studies were performed on basophil granulocytes and their role in allergic reactions (7). Cells, sensitized in vivo or in vitro with ‘reagin’, were challenged with allergen and the basophil response was measured as histamine release. In the early 90s, it was reported that when the basophils empty their histamine containing granulae a structure on the granulae vesicle, CD63, becomes exposed on the cell surface (8). Appearance of CD63 on the surface of basophils, measured with flow cytometry, is now an established procedure to detect allergen or anti-IgE triggering of basophils, and the accessibility of the basophil specific marker CD203c (9) has made it possible to more easily identify basophils.

The objective of this study was to transfuse defined amounts of IgE antibodies to an allergen, and to examine the degree and duration of IgE-reactivity in the recipient, measured by skin prick tests (SPT) and basophil allergen threshold activation (CD-sens). The IgE sensitization characteristics of the basophils were also studied.

Materials and methods

Plasma transfusion recipients

Otherwise healthy, nonallergic patients without IgE antibodies to timothy grass pollen were studied. After admission to the Departments of Orthopedics or Plastic Surgery, Karolinska University Hospital, Stockholm, Sweden for a surgical intervention expected to need plasma transfusion, they were contacted by the Head Anaesthesiologist (NO) and invited for the study. From 2001 to 2004 all in all 10 patients completed the study. Before operation two blood samples, 5 ml without additive and 10 ml with heparin, was taken. If plasma was needed, a standard dose of 2 units each of approximately 300 ml, but containing known concentrations of IgE antibodies to timothy grass pollen (8–205 kU_A/l), selected among the regular plasma donors, (see below) was transfused. Plasma with IgE antibodies to other, potentially harmful allergens, were avoided. The intention was to repeat the blood sampling immediately after the transfusion (‘3 h’ sample), the day after (‘day 1’ sample) and then every second to third day until the basophils became nonresponsive to timothy allergen or no IgE antibodies to timothy allergen could be detected in serum. No transfusions were given during May through September to avoid timothy pollen exposure during the season. No recipient reported any reaction that could be related to the plasma transfusion.

Allergic patients

Patients (n = 20) with asthma and/or rhinoconjunctivitis to grass pollen and with a positive SPT and IgE antibodies to timothy grass, and similar patients (allergic controls) (n = 14) but not sensitized to timothy grass, were chosen for the study. Blood from these patients was collected in heparinized tubes and analyzed for CD63-expression within 5 h. Blood was also collected without additive for quantitation of serum concentration of IgE and IgE antibodies to timothy allergen. The study was performed out of grass pollen season to give the opportunity to measure the chosen allergic parameters in a resting phase. The patients in the study were not on allergy drug treatment and had not been treated with allergen specific immunotherapy, ASIT.

Plasma

At Karolinska 111 consecutive plasma donors were screened for IgE antibodies in serum to common allergens and, if positive, the concentration of IgE antibodies to timothy grass was determined. Plasma from three donors with different levels of IgE antibodies to timothy (8–205 kU_A/l) were chosen and used for the transfusions in this study.

IgE and IgE antibodies

The IgE concentration in plasma and serum was determined using UniCAP^® Total IgE (Pharmacia Diagnostics AB, Uppsala, Sweden) according to manufacturer's instruction. UniCAP^® Specific IgE (Pharmacia Diagnostics AB) was used to determine the concentration of IgE antibodies to a mix of common inhalant (Phadiatop) and food (fx5) allergens and to timothy grass pollen allergen (g6) in plasma donors and transfusion recipients. Analyses were performed according to manufacturer's instruction. The results were expressed in kU_A/l and a positive test was defined as ≥0.35 kU_A/l.

Skin-prick test

The SPT titrations on the allergic patients were performed on the volar side of the forearm with three different concentrations (1000, 10 000 and 100 000 SQU/ml, the latter corresponding to 10 histamine equivalent prick units, HEP) of a commercial timothy extract (ALK-Abello A/S, Hørsholm, Denmark) using a fresh sterile lancet (ALK-Abello A/S). Histamine chloride, 10 mg/ml, (ALK-Abello A/S) was used as a positive control and physiological saline (ALK-Abello A/S) as a negative control. All tests were recorded after 15 min. The wheals were outlined with a fine tipped pen and transferred by transparent tape to an mm paper. The SPT mean wheal area for each allergen dilution was plotted against the corresponding allergen concentration as a dose–response relation. To allow comparison of patients’ sensitivity the timothy allergen dose corresponding to the wheal size of the histamine control was calculated. In order to get a figure that was positively related to the degree of sensitivity and had a numerical value easy to handle this dose value was inverted and multiplied by 10 000. This factor was termed SPT sensitivity (SPT-sens). The higher the value for SPT-sens the higher is the skin allergen sensitivity.

The SPT on three of the plasma recipients were performed, before and 2 days after transfusion, with the 100 000 SQU/ml preparation of the timothy extract (ALK-Abello A/S). A wheal the same size as the histamine control was regarded as a 3+ reaction.

Basophil activation

The CD63 analyses were performed essentially as described (10) with minor modifications. Aliquots of 100 μl blood were incubated 20 min in +37°C with 100 μl RPMI-hepes (negative control) (Gibco Ltd., Paisley, Renfrewshire, UK), fMLP 10⁻⁷ M (positive control) (Sigma Chemical Co., St. Louis, MO, USA), 100 μl anti-IgE 10 μg/ml (positive control) (MIAB, Uppsala, Sweden), 25 μl α-FcɛRI (positive control) (Bühlmann Laboratories, Basel, Switzerland) or 100 μl allergen solution 0.01–1000 SQU/ml (ALK-Abello A/S). After stimulation cells were incubated 25 min on ice in the dark with 18 μl CD203c-PE (Clone: 97A6) (Immunotech, Marseille, France) for basophil identification and 18 μl CD63-FITC (Clone: CLB-gran 12) (Immunotech) for detection of basophil activation. One tube containing blood and RPMI was directly placed on ice and stained for 25 min in the dark with 5 μl α-IgE-FITC (Dako Cytomation, Glostrup, Denmark) and 18 μl CD203c-PE (Immunotech). After antibody staining the red blood cells were haemolyzed by adding 2 ml + 4°C isotonic NH₄Cl-EDTA lyzing solution. The leukocyte suspensions were washed with 2 ml + 4°C PBS. After centrifugation the leukocyte pellet was re-suspended in 300 μl PBS and analyzed in a flow cytometer.

Calculation of IgE per basophils

The number of IgE molecules per basophil was calculated using a FITC-conjugated antibody to IgE (see above) and compared to a standard curve of Calibration Beads (Dako Cytomation). With assigned Molecules of Equivalent Fluorochrome (MEF) values for the fluorescent bead populations these beads enables the transformation of arbitrary units of Mean Fluorescence Intensity (MFI) into absolute units.

Flow cytometric analysis

Leukocytes were analyzed and counted in an EPICS XL flow cytometer (Beckman Coulter Inc., Fullerton, CA, USA). The instrument was calibrated daily with two standardized fluorospheres, Flow-Check (Beckman Coulter) and Flow-set (Beckman Coulter). The percentage of CD63 or IgE-positive cells in the basophil gate (determined by CD203c-staining) was analyzed. The level of CD63-positive basophils in the control samples was below 1.5%. The cut off determining a positive test was set to 3% of CD63-positive basophils.

Definition of CD-sens

Basophil allergen sensitivity was measured as the allergen concentration giving a 50% of maximum CD63% up-regulation. The term CD sensitivity (CD-sens), defined as the inverted value for this threshold allergen concentration multiplied by 100, was used to describe a patient's allergen sensitivity. The higher the value for CD-sens the higher is the basophil allergen sensitivity.

In order to determine the precision of the method the intra- and inter assay variation coefficient was determined using three eight-dose–response curve experiments analyzed by two different persons. An intra assay CV of 13.1% and an inter assay CV of 5.4% was obtained.

Statistical analyses

The Chi-square test was used for statistical analysis of differences in distribution between populations and Spearman rank correlation test and Pearson product-moment correlation for comparison of IgE antibody distributions in different samples. Regression analysis was used for estimation of decrease and confidence limits were estimated using the model. As level of significance a P-values of 0.05 was used.

Results

In vitro assay of allergen sensitivity

When the basophils of a patient allergic to timothy grass were challenged with increasing concentrations of timothy pollen allergen extract a bell-shaped, sometimes with a plateau, dose-response curve of allergen dose and percentage CD63 positive basophils was seen (Fig. 1). The timothy CD-sens, a measure of allergen threshold sensitivity, in the group of 20 patients, had a geometric mean of 20.4 (95% confidence interval: 0.9–438) and the corresponding figures for peak percentage of CD63 positive cells was 54.3 (18.2–100). In contrast, basophils of allergic patients not sensitized to timothy challenged in the same way, did not become significantly activated; a maximum of 1.5% CD63+ basophils was detected.

Figure 1. Two representative dose-response curves after allergen stimulation. () Basophils from an allergic individual and (□) Basophils from patient No. 1 on day 7 after plasma transfusion.

The CD-sens was compared to conventional methods for allergen sensitivity, i.e. SPT-titration and serum IgE antibody quantitation. There was a significant correlation (r = 0.55; P < 0.01) between SPT-sens and CD-sens but not with the percentage of CD63 up-regulation at the allergen concentrations of peak or plateau. The SPT-sens geometric mean was 4.1 (95% confidence interval: 0.1–120). The SPT dose/response curves were reasonably parallel although the variation, expressed as difference in mm² of wheal area between allergen and histamine positive control duplicates, was rather high, CV 52%.

Both the allergic patients’ serum IgE levels (geometric mean 131 kU/l) and the concentrations of IgE antibodies to timothy (geometric mean 8.7 kU_A/l) showed a fairly wide range (95% confidence interval: 13–1355 kU/l and 0.5–157 kU_A/l, respectively). This was also the case for the relative concentration of IgE antibodies, expressed as absolute IgE antibody concentration in percentage of IgE (geometric mean 6.7; 95% confidence interval: 0.6–79). There was a significant correlation (r = 0.76; P < 0.001) between the relative, but not the absolute, concentration of IgE antibodies and CD-sens.

Transfusion of IgE antibody containing plasma

Of the 111 consecutive plasma donors tested 27 (24.3%) had IgE antibodies to a common inhalant or food allergen and 12 (10.8%) were IgE-sensitized to timothy grass pollen allergen (Table 1). When medically indicated, approximately 600 ml of plasma containing low (8 kU_A/l) to high (200 kU_A/l) concentrations of IgE antibodies to timothy, were given during operation. Large amounts of IgE antibodies, i.e. plasma with concentrations >35 kU_A/l, resulted in high concentrations, >3.5 kU_A/l, in serum of the recipient, but also plasma with as low a level as 8 kU_A/l resulted in detectable IgE antibody (Table 2). A significant correlation (r = 0.98; P < 0.001) was found between the amount of IgE antibody given and the observed peak IgE antibody concentration in the patients’‘3 h’ sample. The patient serum IgE antibody levels decreased with a mean T_1/2 of 1.13 days (95% confidence limits 0.35–1.91) (Fig. 2).

Table 1.  Distribution of IgE antibody by concentration among 27/111 plasma donors positive to at least one individual allergen, and 12/111 positive to timothy (g6). At each IgE antibody level interval the number of positive samples, in brackets in percentage of the total number of samples positive to that allergen, is given

IgE-antibody range kUA/l	Any allergen	Timothy (g6)
0.35–1.0	5 (18.5)	3 (25)
1.1–3.5	8 (29.6)	3 (25)
3.6–10.0	7 (25.9)	3 (25)
10.1–35.0	4 (14.8)	2 (16.7)
35.1–100	1 (3.7)	0
>100	2 (7.4)	1 (8.3)
Total No.	27/111 (24.3%)	12/111 (10.8%)

Table 2.  The IgE (kU/l) and IgE antibody, ab, (kU_A/l) concentrations in patient serum before and after transfusion and in transfused plasma (given), and in the ‘3 h sample’, as observed, as expected from body plasma pool distribution and as calculated from observed IgE antibody concentration in ‘day 1 sample’ using a T_1/2 of 1.13 day. The maximal IgE levels were observed already during the very first day after transfusion

Patient	IgE in serum before transfusion	Observed maximal IgE in serum after transfusion	IgE ab concentration in plasma given	Observed IgE ab peak concentration in the ‘3 h’ sample	Expected IgE ab peak concentration in the ‘3 h’ sample	Calculated IgE ab peak concentration in serum
*Two patients, Nos. 3 and 4 had nonreactive basophils before transfusion. NT, not tested.
1	11.8	149	205	33.6	27.3	35.3
2	36.0	621	166	31.6	13.8	37.2
3*	38.8	127	160	24.7	14.2	26.3
4*	NT	88	122	NT	10.6	18.8
5	1.7	98	97.4	22.3	12.3	16.0
6	101	118	42.5	6.0	3.8	4.1
7	96.1	284	42.5	4.8	4.2	3.8
8	6.3	35	18	2.2	1.6	2.3
9	12.0	40	11	1.4	1.1	2.8
10	1.5	40	8	2.7	0.8	1.2

Figure 2. Kinetics of disappearance of IgE antibodies to timothy (g6) from serum of patients transfused with plasma containing varying concentrations of IgE antibody. Mean T_1/2 for IgE antibody in circulation was 1.13 days.

The recipients also had increased concentrations of IgE after the plasma transfusion (Table 2). However, a relation to transfused amounts could not be accurately calculated since sufficient information about blood and plasma units given in relation to the surgical intervention was not available.

Since the ‘3 h’ sample was drawn during the operation it is possible that the patient was not in complete fluid balance. However, there was a significant correlation (r = 0.93; P < 0.001) between observed and expected IgE antibody concentration, the latter calculated from the patient body plasma pool (body weight ×90) (Table 2). There was an equally good correlation (r = 0.98; P < 0.001) with a peak IgE antibody value calculated from observed value in the ‘day 1’ sample using T_1/2 1.13 days. These findings support the notion that there was a balance between intra- and extra vascular distribution of IgE already a few hours after the transfusion.

Passive sensitization of recipients’ mast cells and basophils

Three patients were SPT tested to timothy before operation and found negative. However, on day 2 after the plasma transfusion they all had become 2+ to 3+ positive.

When basophils from the transfused patients were challenged with increasing concentrations of grass pollen allergen the same pattern was seen as in the allergic patients (Fig. 1) and CD-sens could be calculated. The CD-sens was positive already in the ‘3 h’ sample and increased rapidly to reach a peak after 3.4 days (95% confidence interval 1.9–6.1) and then declined over days to weeks (Fig. 3). The geometric mean of the peak CD-sens values for the transfused patients (Table 3) was 7.6 (95% confidence interval: 2.3–25.7), and for the peak percentage of CD63 positive basophils 27.2 (95% confidence interval: 5.2–100). A correlation (r = 0.67; P < 0.03) was found between peak CD-sens, but not percentage of CD63 positive cells, and the peak value of the recipients’ relative IgE antibody concentrations. There was no correlation between the peak CD-sens and percent CD63 positive values.

Figure 3. The CD-sens pattern in transfused patients with a slow (open circles) or a rapid (filled circles) decline. Patients Nos. 2 and 6 have been excluded since no samples were available during the critical time period, i.e. days 1–13 and 1–9, respectively. The straight lines for the functions describing the mean increase (–––––) of the entire material and the T_1/2 for the rapid (- - - - - - - - -) and the slow (−·−·−·−) groups are given.

Table 3.  Observed peak CD-sens values compared to peak percentage of CD63 positive basophils at the optimal allergen concentration (SQU/ml)

Patient	Observed peak CD-sens value	Observed peak (%) CD63+ basophils at an optimal allergen concentration	Optimal allergen concentration used for observed peak (%) CD63+ basophils
1	42	56	100
2	16	45	100
3	49	12	20
4	7	13	100
5	15	93	333
6	3	24	1000
7	0.4	6	1000
8	26	45	50
9	0.5	25	2000
10	12	38	100

Although all transfused patients had a rather homogenous increase in CD-sens (‘T_1/2’ = 1.75 days; 95% confidence interval 1.53–2.12) there seemed to be two patterns in the decrease of the CD-sens activity. One group of four patients showed a very rapid decrease (‘rapid’ group; patients Nos. 3, 4, 9, and 10) with a mean T_1/2 of 4.1 days (95% confidence limits 0.4–7.8 days) while the other group of four patients, had a much slower decrease (‘slow’ group; patients Nos. 1, 5, 7, and 8) with a mean T_1/2 of 10.1 days (95% confidence limits 0–26.4 days) (Fig. 3). Patients Nos. 2 and 6 have been excluded from these calculations since no samples were available during the critical time period, i.e. days 1–13 and 1–9, respectively. Two patients, Nos. 3 and 4, of the ‘rapid’ group had essentially nonreactive basophils before transfusion, the fMLP-, anti-IgE and anti-FcɛRI positive controls resulted in less than 2% up-regulation of CD63.

There was a positive correlation between peak CD-sens value and absolute (r = 0.68; P < 0.02) and relative (r = 0.70; P < 0.02) concentration of IgE antibodies to timothy in transfused plasma. The correlation was most evident for the relative IgE antibody concentration in the ‘rapid group’ (r = 0.98; P < 0.001) but was a significant negative correlation (r = −0.99; P < 0.01) in the ‘slow’ group.

Relation between IgE in serum and on basophils

During the first days after the transfusion there were high serum concentrations of IgE (Table 2, Fig. 4A) and low numbers of IgE molecules per basophil (Fig. 4B), but after the IgE peak the relation was the opposite. There was a correlation between peak serum IgE concentration and peak number of IgE molecules per basophil (r = 0.75; P = 0.05). However, in the very last samples collected, usually about days 8–12, the effect of transfused IgE was not obvious.

Figure 4. A. The distribution of IgE levels (kU/l) in serum of all recipients before (day 0) and after transfusion. B. The number of IgE molecules on basophils per microlitre of blood. After a delay of 1–3 days, when the IgE concentrations were peaking, a plateau was reached followed by a slow decrease due to increasing number of basophils. Patients Nos. 1, 2, 6, and 7 have been excluded since no samples were available during the critical observation period.

The geometric mean of the total number of basophil surface IgE molecules per microlitre of blood, i.e. the product of IgE molecules per cell and basophil count, was before transfusion 80 000 (95% confidence interval: 30 000–2 400 000) and at the peak after transfusion 1 100 000 (95% confidence interval: 160 000–2 800 000). The two recipients (Nos. 3 and 4) with nonreactive basophils preoperative peaked at 1 900 000 and 2 000 000, respectively. The peak number of IgE molecules on basophils from transfused patients was comparable with that of allergic patients, who had a geometric mean of 1 200 000 (95% confidence interval: 330 000–4 000 000).

Discussion

Since the prevalence of IgE-sensitization among blood donors, at least in Sweden, is very high (5), a follow-up of routine plasma transfusions offers a unique way to study passive sensitization. The plasma donors at the Karolinska University Hospital, having high concentrations of IgE antibodies to timothy grass pollen allergens were identified. When their plasma was used as a complement to erythrocytes and electrolyte solutions in cases of severe blood loss, the transfused IgE antibodies to timothy grass as well as their basophil sensitizing capacity was monitored for days to weeks after the transfusion.

The IgE from transfused plasma could be detected in the circulation of the recipient already in the first sample, i.e. 3 h after transfusion. As expected, there was a relation between the amount of IgE antibodies transferred and the concentration observed in the serum of the patient. Studies of the metabolism of IgE, based on injections of radio-labeled monoclonal IgE, has indicated a half life for IgE in circulation in the order of 0.7–4.4 days (11, 12), with 2.3 days being a representative and commonly used figure. Our finding in this study of a mean half-life for IgE of 1.13 days is a bit low compared to previous reports; however, it was obtained using native, antibody-active, IgE.

Different approaches have been used to document the allergen sensitivity of a patient. Provocation challenge tests using the offending substance in food, drug and occupational allergy and allergen extract inhalation tests in nasal and bronchial allergy have been applied (13, 14), but could for ethical and medical reasons not be used in this study. The most commonly used test is the SPT, which is easy to perform, but has a low precision (15). In this study the CV for wheal area duplicates was 52%.

In a few patients in this study a SPT to timothy was performed and found to change from negative before to positive after transfusion. Obviously, the risk of a clinically significant IgE-sensitization of mast cells, through transfusion of plasma containing IgE antibodies, is a reality.

The approach of in vitro challenge of tissue sensitized in vivo is an attractive alternative to in vivo provocation tests. Basophils as target organ were extensively explored before IgE was discovered (7) and allergen sensitivity, calculated by using allergen threshold measurements, was found to give the best information of the patient's allergen sensitivity, although a relation to cell reactivity has occasionally been found (16). The CD63 up-regulation after stimulation at one, high dose of allergen, i.e. basophil cell reactivity, has during recent years been used as a measure of both allergic (17) and nonallergic (18) cell responses. In our study CD-sens, i.e. the allergen concentration giving 50% of maximal up-regulation of CD63 was used as a measure of general allergen sensitivity after transfusion and, in contrast to CD63 up-regulation, it was found to correlate significantly with SPT.

When the recipient's CD-sens was monitored during the first 2 weeks after transfusion, it was found that even transfusion of plasma with a moderate concentration of IgE antibody, e.g. 8 kU_A/l, resulted in allergen reactive basophils. The peak CD-sens of the transfused patients had a geometric mean of 7.6, which is comparable to that of patients with asthma and/or rhinoconjunctivitis due to timothy induced, IgE-mediated allergy (geometric mean 20.4; 95% confidence interval 0.9–438). Thus, passive sensitization of basophils results in recipients as IgE-sensitized as allergic patients.

The allergen sensitivity of the recipient's basophils increased rapidly after the transfusion and the CD-sens peaked at 3.4 days. This time-point is interesting since it might mirror the time-span for basophils to fully mature in the bone marrow before they are released into the circulation. During this period the genes for FcɛRI are turned on, leading to accumulation of receptors on the cell surface in a linear fashion (19), and the density of surface FcɛRI receptors is suggested to be a function of the current IgE concentration. Thus, we propose that the increase in IgE concentration caused by the passive transfer of IgE, leads to a higher FcɛRI expression on the maturing basophils in the bone marrow. Moreover, the composition of surface IgE, as well as IgE antibody specificity, on the newly released basophils mirrors the blend of the IgE from the recipient and the plasma donor, thus giving these basophils the capacity to react to allergen in a similar fashion as basophils from the allergic donor. Since it has previously been reported that basophils circulate in the blood for approximately 1 day (19), one can assume that the main part of the circulating basophils at day 3.4 have matured in a bone marrow environment consisting of a blend of IgE from recipient and donor.

Interestingly, two basophil populations with different CD-sens decline T_1/2 were identified; one with a rapid and one with a slower decline. Earlier studies have shown that one or more mutations in the β-chain of FcɛRI (FcɛRI-β), leading to polymorphism in the receptor, could contribute to more elevated skin test responses to grass and house dust mite, higher concentrations of IgE antibodies to grass and bronchial reactivity to methacholine (20) in the tested individuals. Although not tested in the current study, this observation generates the hypothesis that a current FcɛRI-β polymorphism in individuals contributes to elevated reactions in vitro (basophil response) and in vivo (mast cell response), and to a sustained reactivity.

It has been suggested that the allergen reactivity of basophils is not representative of the tissue localized mast cells. If such a difference is of significant importance in a particular allergic disease, or in certain patients, the use of basophils as a surrogate marker for target organ allergen sensitivity must be questioned. However, much of our information of allergic inflammation in humans comes from studies of basophils, as these are more readily accessible. Limited studies of tissue mast cells have been done, and thus far some of the key features of the basophil translate well to the mast cell. In particular, mast cell and basophil allergen sensitivity appears to be very similar (21).

Obviously, transfusion of even moderate amounts of IgE antibodies represents a potential risk for an allergic reaction. In this study two standard units of plasma, with an IgE antibody concentration in the order of 10–20 kU_A/l, could, as calculated from the T_1/2 of 4 and 10 days, respectively, result in a positive CD-sens up to 2–3 weeks after transfusion in a patient with rapid and 5–7 weeks in one with a slow decline. Regular blood donors with IgE antibodies to common allergens at concentrations >10 kU_A/l are not rare (5). In a recent study, as many as 7.5% of donors in Sweden and 5.6% in Norway had such high IgE antibody levels [unpublished observation], and a few of them to drugs (penicilloyl G), latex and foods (e.g. peanuts).

The mast cell allergen sensitivity seems to disappear even more slowly and studies based on Prausnitz–Küstner titration, indicate a half-life in the order of 2 weeks (22). This difference between mast cells (23) and basophils (24) has been observed also in recent studies on the function of omalizumab. Thus, a recipient will be at risk also when exposed to common inhalant and food allergens outside the hospital environment, many weeks after the transfusion.

In both allergic patients and transfusion recipients there was a relation between concentration of IgE antibodies and CD-sens. Such a relation between IgE antibodies, SPT and food challenge has been observed in other studies, e.g. recently in allergy to peanuts (13). Our findings of a significant correlation between the relative concentration of IgE antibodies and CD-sens are in agreement with studies using a rat mast cell line transfected with the α-chain of the human high-affinity IgE receptor and measuring of cell activation from release of a β-hexosaminidase (25). They reported a good correlation with the relative, but not with the absolute, concentration of IgE antibodies indicating that nonantibody active IgE molecules, when bound to Fcɛ-receptors on the mast cell or basophil surface, can interfere with the ability of the cell to respond to allergen stimulation.

The geometric mean of the total number of basophil surface IgE molecules per microlitre of blood was before transfusion 80 000 and at the peak after transfusion 1 100 000. This peak number of IgE molecules was comparable with that of allergic patients, 1 200 000, showing that passively transferred IgE molecules binds to the basophils to the same extent as in allergic patients. Interestingly, the two recipients (Nos. 3 and 4) with nonreactive basophils pretransfusion did reach peak values for IgE molecules, 1 900 000 and 2 000 000, respectively, and CD-sens, 49 and 7, respectively, as the other although both belonged to the ‘rapid’ decrease group.

In conclusion, transfusing plasma containing IgE antibodies results in a transient sensitization of basophils and mast cells; this reaches the same level as that of allergic patients. There seems to be two populations of basophils; those with a short allergen sensitivity half-life of 4 days and those with a longer half-life, 10 days. Their importance for the expression of allergic disease should be further studied. An IgE-sensitization has the potential to initiate severe transfusion complications during the first days postoperatively but also interfere with the recovery for weeks.

Acknowledgments

We would like to thank Ass. Prof. R. Grönneberg, Dr K. Norrlind and Dr M. Kronquist for help with SPT titrations and BMA I. Ågren-Andersson for performing the ImmunoCAP analyses.