Blood Bank questions

N E W M E T H O D S A N D A P P R O A C H E S

When blood transfusion medicine becomes complicated due to

interference by monoclonal antibody therapy

Marlies Oostendorp,1 Jeroen J. Lammerts van Bueren,2 Parul Doshi,3 Imran Khan,3

Tahamtan Ahmadi,3 Paul W.H.I. Parren,2,4 Wouter W. van Solinge,1 and Karen M.K. De Vooght1

BACKGROUND: Monoclonal antibodies (MoAbs) are

increasingly integrated in the standard of care. The

notion that therapeutic MoAbs can interfere with clinical

laboratory tests is an emerging concern that requires

immediate recognition and the development of

appropriate solutions. Here, we describe that treatment

of multiple myeloma patients with daratumumab, a novel

anti-CD38 MoAb, resulted in false-positive indirect

antiglobulin tests (IATs) for all patients for 2 to 6 months

after infusion. This precluded the correct identification of

irregular blood group antibodies for patients requiring

blood transfusion.

STUDY DESIGN AND METHODS: The IAT was

performed using three- and 11-donor-cell panels.

Interference of daratumumab and three other anti-CD38

MoAbs was studied using fresh-frozen plasma spiked

with different MoAb concentrations. Additionally it was

tested whether two potentially neutralizing agents, anti-

idiotype antibody and recombinant soluble CD38

(sCD38) extracellular domain, were able to inhibit the

interference.

RESULTS: The CD38 MoAbs caused agglutination in

the IAT in a dose-dependent manner. Addition of an

excess of anti-idiotype antibodies or sCD38 protein to the

test abrogated CD38 MoAb interference and successfully

restored irregular antibody screening and identification.

DISCUSSION: CD38 MoAb therapy causes false-

positive results in the IAT. The reliability of the test could

be restored by adding a neutralizing agent against the

CD38 MoAb to the patient’s plasma. This study

emphasizes that during drug development, targeted

therapeutics should be investigated for potential

interference with laboratory tests. Clinical laboratories

should be informed when patients receive MoAb

treatments and matched laboratory tests to prevent

interference should be employed.

D rug interference is a well-known phenomenon

in laboratory medicine, 1

but can be different

for each drug and each analytical method. For

many drugs, interference with laboratory tests

is unknown and is often discovered by chance, for exam-

ple, when unexpected laboratory results are found which

cannot be explained by the patient’s condition.

Monoclonal antibodies (MoAbs) represent a novel

class of therapeutics, which are increasingly used in a

variety of pathologic conditions, including solid tumors,

leukemia, infections, and cardiovascular and inflamma-

tory diseases.2 An important advantage of MoAbs is their

specific targeting. Since many laboratory tests are also

based on specific antibody–antigen interactions, possible

MoAb interference in laboratory medicine is considered

an increasing problem. For example, several MoAbs (sil-

tuximab, rituximab, infliximab, cetuximab, trastuzumab,

bevacizumab, adalimumab, and ofatumumab) were previ-

ously shown to generate false-positive results in serum

protein and immunofixation electrophoresis, tests that are

ABBREVIATIONS: MM 5 multiple myeloma; sCD38 5

soluble CD38; VSB 5 veronal saline buffer.

From the 1 Department of Clinical Chemistry and Haematology,

University Medical Center Utrecht, and 2 Genmab, Utrecht,

The Netherlands; the 3Janssen R&D LLC, Spring House

(Ambler), Pennsylvania; and the Department of

Immunohematology and Blood Transfusion, 4 Leiden University

Medical Center, Leiden, The Netherlands.

This study was funded by Genmab.

Address reprint requests to: Karen M.K. de Vooght, Depart-

ment of Clinical Chemistry and Haematology, University Medi-

cal Center Utrecht, P.O. Box 85500, 3508 GA, Utrecht, The

Netherlands; e-mail: [email protected].

Received for publication August 9, 2014; revision received

March 25, 2015; and accepted April 5, 2015.

doi:10.1111/trf.13150

VC 2015 AABB

TRANSFUSION 2015;55;1555–1562

Volume 55, June 2015 TRANSFUSION 1555

used for diagnosis and follow-up of patients with multiple

myeloma (MM) or Waldenstr€om’s macroglobulinemia.3,4

In a Phase I and II trial with daratumumab, a novel

IgG1j anti-CD38 MoAb which effectively targets and kills

human MM cells,5-7 we observed an unexpected interfer-

ence in routine laboratory tests used in blood transfusion

medicine. All patients receiving daratumumab showed

false-positive indirect antiglobulin tests (IATs), used for

the detection of irregular blood group antibodies.

Although this might only appear to be a clinical laboratory

problem, the interference seriously complicated the selec-

tion of suitable blood products for transfusion for these

patients and was therefore further investigated. Solutions

to prevent MoAb interference were investigated and impli-

cations for patient safety are discussed.

MATERIALS AND METHODS

Additional methods descriptions are provided in the Sup-

porting Information, available in the online version of this

paper.

Study characteristics

MM patients (single center n 5 11, male/female 5 7/4,

age 58 6 9 years) were enrolled in a Phase I and II safety

and dose escalation study with daratumumab (HuMax-

CD38, Genmab A/S, Copenhagen, Denmark; Clinical Trial

Identifier NCT00574288, http://clinicaltrials.gov/show/

NCT00574288). The clinical trial was approved by the

institutional ethics committee and written informed con-

sent was obtained from all patients.

Patients received a low dose of daratumumab 1 day

before the first full dose of 8 to 16 mg/kg. Three weeks

after the first full dose, patients received another dose.

The next day, 8 to 16 mg/kg daratumumab was given in a

weekly interval for 8 weeks. Peak daratumumab concen-

trations in serum were more than 100 mg/mL for all patients (range, 110-438 mg/mL). As the multicenter trial is still ongoing, follow-up times for patients regarding the

data on required blood transfusions are variable.

(In)direct antiglobulin testing

To investigate the ability of daratumumab to induce in

vitro red blood cell (RBC) agglutination, fresh-frozen

plasma (FFP) was spiked with 0.01, 0.1, 1.0, and 10.0 mg/ mL daratumumab. IAT was subsequently performed in

the low-ionic-strength solution (LISS) gel column aggluti-

nation technique with anti-IgG present in the gel matrix,

using a three-cell Surgiscreen panel or an 11-cell Resolve

C panel, both containing 0.8% donor RBC suspensions (all

reagents from Ortho Clinical Diagnostics, Raritan, NJ). As

a control, the IAT was repeated with RBCs from MM

patients not receiving daratumumab. For the direct anti-

globulin test (DAT), a 0.8% suspension of the patient’s

own RBCs was made in LISS diluent (Bio-Rad, Hercules,

CA). This was subsequently tested in the LISS/Coombs gel

column technique (Bio-Rad), containing polyspecific anti-

IgG and anti-C3d within the gel matrix. Autocontrol

experiments were performed in the LISS gel agglutination

column technique, by mixing a 0.8% suspension of the

patient’s own RBCs with the patient’s plasma. All aggluti-

nation strengths were graded from 0 to 41 (0 5 no aggluti-

nation; 0.51 5 very weak agglutination; 11 5 weak

agglutination; 21 5 agglutination; 31 5 strong agglutina-

tion; 41 5 very strong agglutination).

Antibody elution from RBCs

Antibodies were recovered from RBCs by acid elution

using an elution kit (Gamma Elu Kit II, Immucor Inc.,

Norcross, GA) according to the manufacturer’s instruc-

tions. In brief, RBCs were washed four times with wash

buffer as provided by the manufacturer. Next, washed

RBCs were incubated with the eluate solution for approxi-

mately 30 seconds at room temperature. After centrifuga-

tion and correction of the pH to 6.4 to 7.6, the obtained

eluate was used in an IAT as described earlier.

Agglutination with other CD38 antibodies

To test whether in vitro RBC agglutination is a class-

specific issue, three other CD38 antibodies were pro-

duced: Clones 38SB19, MOR03087, and Ab79. Clones

38SB19 and MOR03087 represent surrogates for the anti-

CD38 SAR650984 (humanized) and MOR202 (human),

respectively. Clone Ab79 is a human anti-CD38 in preclini-

cal development. The heavy- and light-chain sequences of

38SB19, MOR03087, and Ab79 were obtained from patent

applications WO 2008/047242, WO 2012/041800, and WO

2012/092612, respectively, and cloned into mammalian

expression plasmids containing human j and c1 constant regions. The antibodies were generated by transient trans-

fection in HEK293 cells as described by Vink and col-

leagues. 8

IAT was performed using FFP spiked with 0.01,

0.1, 1.0, and 10.0 mg/mL antibodies, as described above.

Preventing MoAb interference using anti-idiotype

antibodies and soluble CD38

The prevention of anti-CD38 MoAb interference was stud-

ied by repeating the indirect antiglobulin experiment using

FFP spiked with 10.0 mg/mL daratumumab and adding a neutralizing daratumumab anti-idiotype antibody (see

below) at five and 10 times the daratumumab concentra-

tion. The anti-idiotype antibody was also tested using

plasma of MM patients participating in the current trial

(i.e., daratumumab present in vivo and not added in vitro).

The performance of the anti-idiotype antibody was

further investigated by spiking plasma of a patient

with known irregular antibodies (anti-E and anti-K) with

10 mg/mL daratumumab or the combination of

OOSTENDORP ET AL.

1556 TRANSFUSION Volume 55, June 2015

daratumumab and a five- or 10-fold excess of anti-

idiotype antibody. Antibody identification experiments

were subsequently performed using an 11-cell screening

panel.

Recombinant soluble CD38 (sCD38) was investigated

as another potential solution to prevent MoAb interfer-

ence (see below). To this extent, sCD38 was added to

plasma spiked with 10 mg/mL daratumumab or 38SB19 in 10- and 20-fold higher concentrations (concentration dif-

ference with anti-idiotype antibody due to the mono- and

bivalent binding capacity of sCD38 and anti-idiotype,

respectively). Hereafter, standard IATs were performed

using a 3-cell screening panel. The effect of sCD38 was

subsequently evaluated using plasma of a patient with

known anti-K spiked with daratumumab.

Generation of the daratumumab anti-idiotype

antibody

Anti-idiotype antibodies against daratumumab were

generated by BioGenes (Berlin, Germany). Briefly, 8-

week-old female BALB/C mice (Charles River Laborato-

ries, Sulzfeld, Germany) were immunized with daratu-

mumab. After isolation of mouse splenocytes and fusion

with SP2/0 mouse myeloma cells (DSMZ, Braunschweig,

Germany), the resulting hybridomas were tested for

binding to daratumumab by an enzyme-linked immuno-

sorbent assay (ELISA). Binding to the human MoAb

HuMab-KLH, a human IgG1 antibody directed against

mariculture keyhole limpet hemocyanin (KLH), was

used in the ELISA for negative selection. 9

Positive clones

were selected and stable antibody-producing clones

were generated by two rounds of limiting dilution clon-

ing. The generated anti-daratumumab clones were

tested for their potential to block daratumumab binding

to CD38-expressing cells. Anti-idiotype Clone 5-3-9 of

the mouse IgG1j subclass was selected for its potency

to block the interaction between daratumumab and

CD38.

Cloning, expression, and purification of sCD38

The sCD38 was generated by transient transfection in

HEK293 cells as described by de Weers and colleagues.5 A

construct similar to the previously described pEE13.4-

HACD38 was made synthetically and was fully codon

optimized (GeneArt, Regensburg, Germany), replacing the

HA tag encoding part by a His tag (HHHHHH) encoding

part. The construct was cloned in pEE13.4 and named

pEE13.4HisCD38. Plasmid DNA was transiently trans-

fected in HEK293F cells using 293fectin (both Invitrogen,

Carlsbad, CA). Proteins were purified from culture super-

natant by chromatography (BD Talon, BD Biosciences,

Palo Alto, CA), and their appropriate molecular weights

were confirmed by sodium dodecyl sulfate-

polyacrylamide gel electrophoresis.

Antibody-induced complement-dependent

cytotoxicity of human RBCs

The assay was performed with whole blood from three

healthy donors collected in heparin tubes. The number of

RBCs was determined after counting in the presence of try-

pan blue, after which the RBCs were washed with RPMI

1640. Finally, the RBCs were diluted to 1 3 10 8

cells/mL in

veronal saline buffer (VSB11, Lonza, Basel, Switzerland).

Test antibodies were diluted 23 in VSB11, added to RBCs,

and incubated for 30 minutes at 48C. After being washed

twice with VSB11, cells were resuspended in VSB11 and

active or inactivated normal human serum (inactivated for

30 min at 608C) was added. As a control for 100% lysis, water

was added to the RBCs. Cells were incubated for 1 hour at

378C. Subsequently, free hemoglobin (Hb) was measured in

the supernatant using a Hb assay kit (Abnova, Taipei City,

Taiwan) according to the manufacturer’s instructions.

RESULTS

Daratumumab infusion causes positive irregular

antibody screening results

Regular blood group serologic testing (i.e., irregular anti-

body screening) is standard of care for all hematologic

patients in the University Medical Center Utrecht, even

when there is no direct clinical need for transfusion. This

is a precautionary measure to allow quick delivery of RBCs

if requested and is performed using direct and IATs. Before

daratumumab treatment, plasma of all patients showed

negative direct and IATs. However, after MoAb infusion,

positive results with generally 21 reactions strengths were

found for the IAT for all patients, suggesting interference

of daratumumab in the antiglobulin test (see Fig. S1, avail-

able as Supporting Information in the online version of

this paper, for a graphical representation on the mecha-

nism of the MoAb interference). Results remained positive

for 2 to 6 months after the last daratumumab infusion and

were not only observed in the gel column technique, but

also in the tube technique using albumin or polyethylene

glycol (not shown). Interestingly the DAT was negative

after infusion for all patients (n 5 11, multiple tests per

patient), indicating that there are no IgGs bound to the

RBCs of daratumumab-treated patients. In addition, the

IAT autocontrol, which tests the agglutination of the

patients’ plasma with their own RBCs, was also negative

for all patients. This implies that daratumumab present in

the patient’s plasma does not induce agglutination of the

patient’s own RBCs in the IAT. These data were confirmed

by the observation that acid eluates prepared from

daratumumab-treated patients’ RBCs did not show any

agglutination with the patients’ own RBCs as well as donor

RBCs (n 5 11). Taken together, these results suggest a

rapid in vivo clearance of a small RBC fraction to which

daratumumab is bound. This is supported by a minor, but

TRANSFUSION COMPLICATED DUE TO MoAb THERAPY

Volume 55, June 2015 TRANSFUSION 1557

clinically nonsignificant, decrease in Hb levels after infu-

sion and an increase in reticulocyte count (Fig. 1).

The ability of daratumumab to indirectly induce RBC

agglutination in vitro was further investigated by repeating

the IAT using FFP spiked with increasing doses of daratu-

mumab. As shown in Table 1, daratumumab induced RBC

agglutination in a dose-dependent manner. No differences

were found in agglutination patterns when using RBCs

from untreated MM patients (Table 1).

False-positive IATs are class-specific for anti-CD38

MoAbs

IAT was repeated with plasma spiked with the humanized

anti-CD38 MoAb 38SB19 and human anti-CD38 MoAbs

MOR03087 and Ab79. Comparable dose-dependent agglu-

tination patterns were found as described for daratumu-

mab (Table 1), which were of the same magnitude for

38SB19 and somewhat weaker for MOR03087 and Ab79

(Table S1, available as Supporting Information in the

online version of this paper). This indicates that the false

positive IAT is not unique for daratumumab, but is a

class-specific problem for anti-CD38.

Daratumumab can be recovered from donor RBCs

by acid elution

Daratumumab was recovered from donor RBCs incubated

with daratumumab-spiked plasma using acid elution. The

eluate contained sufficient daratumumab to again induce

weak agglutination of donor RBCs in the IAT (Table S2,

available as Supporting Information in the online version

of this paper). The mean daratumumab concentration in

the eluate was 86.8 6 36.7 ng/mL, as measured using an

ELISA. This indicates that daratumumab binding to RBCs

is indeed the cause of RBC agglutination in the IAT.

CD38 shows no age-dependent expression pattern

We investigated whether expression of CD38 depends on

RBC age. Therefore, RBCs were separated into five age

fractions using discontinuous Percoll gradient centrifuga-

tion. All fractions were subsequently used for IAT with

daratumumab-spiked plasma. No differences were

observed in agglutination patterns between the RBC frac-

tions (Table S3, available as Supporting Information in the

Fig. 1. Infusion of daratumumab (dashed vertical lines)

resulted in a small, clinically nonsignificant decrease in Hb

levels (A) and a compensatory rise in reticulocyte count (B).

Adverse events encountered with daratumumab infusion did

not include anemia or hemolysis and patients did not

require blood transfusion.

TABLE 1. RBC agglutination patterns of plasma supplemented with daratumumab (top) and another CD38 antibody 38SB19 (bottom) in the IAT using a three-cell RBC screening panel or RBCs from untreated MM patients

Cell MM patient

Anti-CD38 1 2 3 1 2

Daratumumab (mg/mL) 0.00 – – – – – 0.01 – – – ND ND 0.1 0.51 0.51 0.51 ND ND 1.0 11 11 11 11 21 10 21 21 21 11 21

38SB19 (mg/mL) 0.00 – – – ND ND 0.01 – – – ND ND 0.1 0.51 0.51 0.51 ND ND 1.0 11 11 11 ND ND 10 11 11 11 ND ND

ND 5 not determined.

OOSTENDORP ET AL.

1558 TRANSFUSION Volume 55, June 2015

online version of this paper), although the reticulocyte

fraction could not be clearly evaluated due to the absence

of significant amounts of reticulocytes in healthy adults.

Nevertheless, CD38 expression on RBCs does not appear

to be restricted to certain cellular ages.

Flow cytometry analysis revealed a limited level

of staining by daratumumab of CD38 molecules on

reticulocytes and RBCs (see Fig. S3, available as Support-

ing Information in the online version of this paper), which

corresponds to previously published results.10 It is likely

that CD38 is present on all RBCs, albeit at a different den-

sity per cell. Consequently, only a small number of RBCs

has sufficient CD38 density to allow relevant levels of dar-

atumumab binding, resulting in in vivo clearance or in

vitro interference in the IAT.

Daratumumab-induced RBC depletion is not

caused by complement-mediated lysis

After daratumumab infusion, patients showed a Hb

decrease of approximately 1.6 g/dL (Fig. 1A). In vitro

experiments did not show daratumumab-induced com-

plement-mediated lysis (Fig. 2), suggesting that

complement-mediated lysis is not involved in the clear-

ance of daratumumab-loaded RBCs. We therefore specu-

late that the small daratumumab-loaded RBC fraction

disappears from the circulation by Fc-receptor–mediated

clearance in the spleen. 11

Blocking the interference of CD38 MoAbs in the

IAT

We investigated whether a specific daratumumab anti-

idiotype antibody was able to abrogate daratumumab-

mediated RBC agglutination in the IAT. RBC agglutination

induced by plasma spiked with 10 mg/mL daratumumab was completely blocked using daratumumab anti-

idiotype antibodies at five- and tenfold excess concentra-

tions (Table 2). The anti-idiotype antibody was also tested

using plasma from MM patients who were treated with

daratumumab. Addition of anti-idiotype antibodies in the

laboratory assay prevented agglutination in the IAT (Table

2). In Fig. S2, available as Supporting Information in the

online version of this paper, a graphical representation of

Fig. 2. Daratumumab-mediated complement-dependent

cytotoxicity (CDC) was evaluated in three different donors.

No significant Hb release was observed when RBCs were

incubated with daratumumab in the presence of 10% active

normal human serum (NHS), indicating that daratumumab

does not induce complement mediated-lysis of RBCs. Anti-P

was used as positive control for CDC lysis. Water was added

to the RBCs as a control for 100% lysis. Results are expressed

as mean 6 SD, n 5 3. (w) Active NHS; (�) inactivated NHS.

TABLE 2. False-positive irregular antibody screening results can be effectively blocked using the daratumumab anti-idiotype antibody at a five- or 10-fold excess concentration (top). The anti-idiotype antibody also successfully

diminishes positive reactions caused by daratumumab present in plasma of a daratumumab-treated patient (in vivo daratumumab concentration > 200 mg/mL; middle). sCD38 extracellular domain protein (sCD38) efficiently prevents

the interference of both daratumumab and 38SB19 (bottom)

Cell 1 Cell 2 Cell 3

Plasma 1 10 mg/mL dara 11 11 11 Plasma 1 10 mg/mL dara 1 53 anti-idiotype – – – Plasma 1 10 mg/mL dara 1 103 anti-idiotype – – – Plasma 1 10 mg/mL dara, corrected for dilution 11 11 11

Dara patient plasma (>200 mg/mL dara) 21 21 21 Dara patient plasma 1 53 anti-idiotype – – –

Plasma 1 10 mg/mL dara 11 11 11 Plasma 1 10 mg/mL dara 1 103 sCD38 – – – Plasma 1 10 mg/mL dara 1 203 sCD38 – – – Plasma 1 10 mg/mL dara, corrected for dilution 11 11 11 Plasma 1 10 mg/mL 38SB19 11 11 11 Plasma 1 10 mg/mL 38SB19 1 103 sCD38 – – – Plasma 1 10 mg/mL 38SB19 1 203 sCD38 – – – Plasma 1 10 mg/mL 38SB19, corrected for dilution 11 11 11

TRANSFUSION COMPLICATED DUE TO MoAb THERAPY

Volume 55, June 2015 TRANSFUSION 1559

the mechanism by which anti-idiotype antibodies prevent

daratumumab-induced RBC agglutination is provided.

Next, the performance of the anti-idiotype antibody

was tested using daratumumab-spiked plasma from a

randomly selected subject with known anti-E and anti-K

antibodies. As expected, daratumumab caused agglutina-

tion of all RBC suspensions of the 11-cell identification

panel and the donor’s own RBCs (Table 3). Adding the

anti-idiotype antibody in a fivefold excess concentration

resulted in the original agglutination pattern (i.e., without

daratumumab), typical for the presence of anti-E and

anti-K (Table 3). This indicates that the anti-idiotype anti-

body does not interfere with the binding of clinically rele-

vant irregular antibodies and allows correct irregular

antibody identification.

As an alternative to a daratumumab-specific anti-

idiotype antibody, sCD38 extracellular domain protein

(sCD38) was tested as a generic solution to prevent inter-

ference by anti-CD38 MoAbs. As shown in Table 2, sCD38

can be successfully applied to block interference by dara-

tumumab as well as 38SB19. sCD38 also allowed correct

identification of known irregular antibodies in plasma

spiked with daratumumab (not shown). sCD38 therefore

provides a generic solution to prevent false-positive indi-

rect antiglobulin results caused by anti-CD38 MoAbs.

DISCUSSION

Present findings

MoAbs are a rapidly expanding class of drugs with

increasing clinical applications. The possible interference

of such therapeutics in laboratory testing, however, is

often poorly investigated. Here, we describe that infusion

of the monoclonal anti-CD38 daratumumab causes a

false-positive result in the IAT used in blood transfusion

medicine. We found that a small fraction of RBCs express

a low level of CD38 molecules per cell, which appears

unrelated to RBC age. In all patients, daratumumab infu-

sion resulted in a mild and temporal decrease in Hb,

accompanied by an increase in reticulocyte count, with-

out resulting in clinically relevant anemia (Fig. 1). We

speculate that this decrease in Hb is likely not due to

complement-mediated lysis (Fig. 2), but due to Fc-

receptor–mediated clearance in the spleen. 11

It was fur-

thermore observed that anti-CD38 MoAb interference in

the IAT is not specific for daratumumab, as comparable

dose-dependent interference was also observed for three

additional anti-CD38.

Clinical perspective

The use of daratumumab leads to in vitro RBC agglutina-

tion and thereby to false-positive results in the IAT, which

is used to detect irregular antibodies. Although this may

appear only a clinical laboratory problem, there are

important consequences for blood transfusion medicine,

as the presence of irregular antibodies to clinically rele-

vant blood groups cannot be ruled out using the standard

tests. This concern should be recognized when patients

require a blood transfusion. In Phase I and II trials in

which 10 of 78 MM patients treated with daratumumab

worldwide required transfusion, no major transfusion-

related events were observed. It should be noted that

these transfusions were not directly related to the small

Hb decrease caused by daratumumab, but were due to

the underlying hematologic malignancy or a completely

unrelated condition or therapy (e.g., hip replacement sur-

gery). All patients were required to undergo blood typing

before being treated with daratumumab. In addition,

potential mitigation strategies are under development,

that can be implemented across blood banks globally to

prevent any potential blood transfusion problems, two of

which (i.e., the anti-idiotype antibody and sCD38) are

described in the present work.

Mitigation strategies

Different scenarios on how to cope with MoAb interference

can be envisioned depending on the clinical condition of

the patient. During acute, life-threatening situations non–

cross-matched blood group O D– RBCs can be transfused

as this product is suitable for any combination of the ABO

and D blood types. This strategy, that doesn’t take the

potential presence of alloantibodies into account, is identi-

cal for patients not receiving MoAb therapy. In elective sit-

uations, extensive typing and matching for clinically

TABLE 3. RBC agglutination patterns of an 11-cell identification panel with plasma from a patient with known irregular antibodies against blood groups E and K and spiked with daratumumab. Cells 3 and 6 of the identification panel were E1 and Cells 2 and 7 were K1. Adding a fivefold excess daratumumab anti-idiotype antibody recovers

the original agglutination pattern and allows correct identification of the known irregular antibodies

Cell

1 2 3 4 5 6 7 8 9 10 11 Autocontrol

Plasma – 31 31 – – 21 31 – – – – – Plasma 1 dara 11 31 31 0.51 11 31 31 11 11 11 21 21 Plasma 1 dara 1

anti-idiotype – 31 31 – – 21 31 – – – – –

OOSTENDORP ET AL.

1560 TRANSFUSION Volume 55, June 2015

relevant blood group antigens (i.e., D, C, c, E, and e and

Kell, Kidd, Duffy, and MNS antigens) can be performed.

Although this strategy prevents mismatching for the most

common blood groups and also prevents development of

irregular antibodies against these blood groups, it has sev-

eral disadvantages. First, it is very time-consuming. Sec-

ond, only a limited number of matching donors will be

available, likely resulting in shortage of compatible blood

products if the blood loss is too extensive. Third, and most

importantly, the presence of other irregular antibodies still

cannot be excluded due the positive cross-matching results

caused by the anti-CD38 MoAb. Although posttransfusion

alloimmunization occurs in only 2% to 3% of the general

population,12 the incidence increases to approximately 9%

in patients with hematologic malignancies.13 Alloantibod-

ies can be directed against any of over 300 different blood

groups and can cause (delayed) hemolytic transfusion

reactions if they remain undetected. This risk can be easily

avoided if the MoAb interference is blocked during labora-

tory testing.

We developed two solutions to overcome the interfer-

ence of anti-CD38 MoAbs in the IAT. First, a specific neu-

tralizing anti-idiotype antibody, a reagent that is usually

generated during drug development programs, can be

employed and, second, the recombinant sCD38 extracel-

lular domain, which could provide a generic solution to

attenuate the interference of anti-CD38. Both approaches

resulted in abrogation of the interference of anti-CD38 in

the IAT, without interfering with irregular antibody detec-

tion. It is therefore recommended that national reference

laboratories for blood transfusion medicine are provided

with these reagents, to allow safe and timely blood trans-

fusion for patients receiving anti-CD38 MoAb therapy. In

addition, patients may carry a blood transfusion card indi-

cating that they receive anti-CD38 MoAb therapy.

MoAb interference in other clinical laboratory tests

The described interference with blood group serologic

testing represents an example of the type of interactions

that may occur between biologics present in the patient’s

serum and specific tests in the clinical laboratory. Such

interference may occur more often with the increasing

use of MoAb therapies and may apply to a wide variety of

laboratory tests, as also previously demonstrated for sev-

eral MoAbs in serum protein and immunofixation electro-

phoresis.3 Specifically, potential interference with

laboratory tests that are based on selective antibody–anti-

gen interactions, similar to the ones targeted by the thera-

peutic molecule, should be investigated during drug

development. A close interaction between the researchers,

clinicians, and clinical laboratory experts is therefore

critical.

In conclusion, the potential interference of MoAb

therapeutics with laboratory tests is considered an

increasing problem due to their increasing clinical use. As

described here, interference of anti-CD38 MoAbs in the

IAT may delay or even prevent the selection of suitable

blood products for transfusion. Proper solutions, like the

two solutions presented in this work, should therefore be

developed in concordance with the development of the

MoAb therapeutic, to allow for convenient and correct

laboratory testing and to ensure patient safety. As these

solutions are to be used for patient follow-up during

MoAb treatment, we suggest an extension of the definition

of companion diagnostics, to also include these specific

reagents which prevent inference of MoAbs in laboratory

testing. Finally, as the type of interference of the therapeu-

tic with clinical laboratory testing is not always predict-

able, it is advised that clinical laboratories should always

be informed when patients are treated with biologics.

ACKNOWLEDGMENTS

The authors thank Gerdien Walbeek and Brigitte van Oirschot

(Laboratory of Clinical Chemistry and Haematology, UMC

Utrecht, The Netherlands) for performing the blood group sero-

logic testing and the Percoll age separation of RBCs, respectively.

MO, JJLvB, PWHIP, WWvS, and KMKdV were responsible for writ-

ing the manuscript; literature search; figure design; study design;

data collection, analysis, and interpretation; and final approval of

manuscript. PD, IK, and TA were responsible for writing the

manuscript, data interpretation, and final approval of

manuscript.

CONFLICT OF INTEREST

MO, WWvS, and KMKdV have disclosed no conflicts of interests.

JJLvB and PWHIP are employees of Genmab, own Genmab war-

rants and/or stock, and are listed as inventors on daratumumab

patent applications owned by Genmab. Genmab funded this

study. PD, IK, and TA are employees of Janssen. Janssen has an

exclusive worldwide daratumumab license and development

agreement from Genmab.

REFERENCES

1. Young D. Effects of drugs on clinical laboratory tests. 5th ed.

Washington, DC: AACC Press; 2000.

2. Ruuls SR, Lammerts van Bueren JJ, van de Winkel JG, et al.

Novel human antibody therapeutics: the age of the Umabs.

Biotechnol J 2008;3:1157-71.

3. Genzen JR, Kawaguchi KR, Furman RR. Detection of a

monoclonal antibody therapy (ofatumumab) by serum pro-

tein and immunofixation electrophoresis. Br J Haematol

2011;155:123-5.

4. McCudden CR, Voorhees PM, Hainsworth SA, et al. Interfer-

ence of monoclonal antibody therapies with serum protein

electrophoresis tests. Clin Chem 2010;56:1897-9.

TRANSFUSION COMPLICATED DUE TO MoAb THERAPY

Volume 55, June 2015 TRANSFUSION 1561

5. de Weers M, Tai YT, van der Veer MS, et al. Daratumumab, a

novel therapeutic human CD38 monoclonal antibody, indu-

ces killing of multiple myeloma and other hematological

tumors. J Immunol 2011;186:1840-8.

6. van der Veer MS, de Weers M, van Kessel B, et al. The thera-

peutic human CD38 antibody daratumumab improves the

anti-myeloma effect of newly emerging multi-drug thera-

pies. Blood Cancer J 2011;1:e41.

7. van der Veer MS, de Weers M, van Kessel B, et al. Towards

effective immunotherapy of myeloma: enhanced elimination

of myeloma cells by combination of lenalidomide with the

human CD38 monoclonal antibody daratumumab. Haema-

tologica 2011;96:284-90.

8. Vink T, Oudshoorn-Dickmann M, Roza M, et al. A simple,

robust and highly efficient transient expression system for

producing antibodies. Methods 2014;65:5-10.

9. Lammerts van Bueren JJ, Bleeker WK, Bøgh HO, et al. Effect

of target dynamics on pharmacokinetics of a novel thera-

peutic antibody against the epidermal growth factor recep-

tor: implications for the mechanisms of action. Cancer Res

2006;66:7630-8.

10. Albeniz I, Demir O, T€urker-Sener L, et al. Erythrocyte CD38

as a prognostic marker in cancer. Hematology 2007;12:409-

14.

11. Frank MM, Schreiber AD, Atkinson JP, et al. NIH conference.

Pathophysiology of immune hemolytic anemia. Ann Intern

Med 1977;87:210-22.

12. Heddle NM, Soutar RL, O’Hoski PL, et al. A prospective

study to determine the frequency and clinical significance of

alloimmunization post-transfusion. Br J Haematol 1995;91:

1000-5.

13. Schonewille H, Haak HL, van Zijl AM. Alloimmunization

after blood transfusion in patients with hematologic and

oncologic diseases. Transfusion 1999;39:763-71.

SUPPORTING INFORMATION

Additional Supporting Information may be found in the

online version of this article at the publisher’s website:

Table S1. RBC agglutination patterns of fresh frozen

plasma spiked with two other anti-CD38 antibodies

MOR03087 (top) and Ab79 (bottom) in the indirect anti-

globulin test using a 3-cell RBC screening panel.

Table S2. Agglutination strengths of RBC eluates pre-

pared from donor RBCs incubated with daratumumab-

spiked plasma using a 3-cell screening panel. The aver-

age daratumumab concentration in the eluate was

86.8 6 36.7 ng/mL, as measured using an ELISA. No

human IgG was detected in acid eluate control samples

of RBCs that were not incubated with daratumumab.

Table S3. Agglutination patterns of age-separated RBCs

from three healthy donors with daratumumab-spiked

plasma. For Fraction 1, agglutination patterns could not

be determined due to a very low yield, likely caused by

the absence of significant amounts of reticulocytes in

healthy donors.

Fig. S1. Daratumumab (DARA) in the patient’s serum

binds to the test RBCs. After adding the anti-IgG rea-

gent, RBC agglutination is observed, thereby generating

a false positive result. The presence of irregular antibod-

ies is masked by the presence of DARA.

Fig. S2. Daratumumab (DARA) specific anti-idiotype

antibodies are added to the patient’s serum and bind to

daratumumab. If the patient has no irregular antibod-

ies, the anti-idiotype blocks the binding of DARA to

RBCs and no agglutination is observed in the indirect

antiglobulin test. However, if the patient does have

irregular antibodies, the anti-idiotype also specifically

binds DARA and RBC agglutination is solely due to the

presence of irregular blood group antibodies. Inhibition

of DARA binding to RBCs was also obtained using a

sCD38 protein.

Fig. S3. A) Dot-plot of double staining, showing the

erythrocyte population in green (CD235a-FITC1/CD71-

APC-) and the reticulocyte population in red (CD235a-

FITC1/CD71-APC1). B) Histograms showing the PE-

fluorescence for reticulocytes and erythrocytes. Results

indicate that red blood cells express low levels of CD38.

OOSTENDORP ET AL.

1562 TRANSFUSION Volume 55, June 2015