Evaluation of non-specific reactivity rates across three immunohematology testing platforms with assessment of subsequent specific alloantibody development

Abstract

Background

Gel and solid phase testing platforms are commonly used for pretransfusion testing but are also prone to nonspecific reactivities (NSP) compared to tube testing. This study aims to characterize the features of NSPs in these platforms and how frequently NSPs precede specific antibody detection.

Methods

All antibody screens and identification panels performed between 5/2021–4/2023 by two automated and one manual platforms at a single institution were analyzed with IRB approval: solid-phase (Immucor, Norcross, GA), automated gel (Grifols, Los Angeles, CA) and manual gel (Ortho Diagnostics, Raritan, NJ). Testing platform, antibody identification, strength of reactivity, number of reactive reagent red cells, and RBC antibody history were extracted from chart review. Differences in reactivity strength were compared using the Kruskal-Wallis test, and differences in the number of cells used to establish diagnoses of NSPs were compared using two-way ANOVA.

Results

During the 24-month study period, a total of 156,247 antibody screens were performed with 5,491 positive screens, and NSP identified on 358/5491 panels (6.6%). 271/5491 (4.9%) panels had NSP detected on non-tube platforms: 95 on solid-phase, 123 on automated gel, and 53 on manual gel. The mean strength of reactivity by agglutination for NSP was 1.7, 0.5, and 1.4 respectively (p<0.001). In addition, 9/271 (3.3%) of all panels with NSP on gel or solid-phase were associated with detection of specific alloantibodies in subsequent panels.

Conclusion

Nonspecific reactions differ by platforms in specificity and strength. Solid-phase shows stronger reactivity while automated gel has higher rates of positivity.

Introduction

In recent years, there has been a significant shift from traditional bench testing to automated testing in the field of transfusion medicine.^1–3 Automated platforms, such as solid-phase technology and gel columns, have become routine for pretransfusion testing. This transition has not only addressed the increased demand for testing but also enabled the identification of some alloantibodies that were previously undetectable with bench methods due to enhanced sensitivity.⁴ However, the same features that enhance antibody detection also introduce an increased risk of nonspecific reactivities (NSP). Studies have reported NSP rates ranging from 13.8% to 29.8% in positive antibody panels, with variations attributed to differences in tested populations, sample sizes, and testing platforms.^5–7 To address NSPs, various methods have been employed to determine whether these reactions represent newly forming alloantibodies or those just below the limit of detection for specific identification.⁸ These methods include specialized testing for low incidence antigens and the application of enzymes.^8–12 Yet another possible cause of NSPs may be chemicals or drugs that may interact with reagents, representing false-positive results.¹³

This study aims to characterize the properties of nonspecific antibodies across solid-phase and two gel-based platforms. A major advantage of this study is pretransfusion testing performed with all three platforms were used by the same group of staff and technologists at a single site. By comparing these platforms in terms of NSP frequency and reactivity strength at a single site, the study seeks to develop strategies for managing NSPs and to assess how often NSPs precede the detection of specific antibodies on these platforms.

Materials and Methods

Data Collection

All antibody screens and identification panels conducted at a large academic medical center between May 2021 and April 2023 were included in this analysis. This study and the accompanying chart review was approved by the Yale University Institutional Review Board. Antibody identification panels from the study period were reviewed twice by the research team, and all panels exhibiting nonspecific reactivity were further examined and archived on a cloud-based platform. Data collected included testing platform, ABO-Rh results, antibody screen and identification results, agglutination reactivity strength for each reagent cell, direct antiglobulin test results, elution results, and cold screen results. Additionally, transfusion history, pregnancy history, past medical history, and RBC antibody history were extracted from the electronic medical record and laboratory information system.

For panels with NSP, reactivity strength was categorized as 0, weak+ (0.5), 1+, 2+, 3+, or 4+. When NSP was accompanied by an alloantibody, only reactive cells not corresponding to the specific alloantibody were included in the analysis. Similarly, in cases involving Rh immune globulin (RHIg) administration, only D-negative cells were studied.

NSPs were assessed using multiple platforms with the same sample to differentiate NSPs from reactions associated with specific alloantibodies. In these instances, only the panel showing NSP was included in the study, while negative panels were noted only if an alloantibody was subsequently identified.

Panels with NSPs for patients with a history of alloantibodies—whether present prior to or identified after the NSP—were subjected to additional evaluation. NSP reagent cells were further analyzed for positivity/negativity and homozygosity/heterozygosity for the relevant antigens. An association between NSP and alloantibody formation was investigated by examining the proportion of NSP cells positive for the antigen associated with the alloantibody.

Pretransfusion Testing

ABO and RhD typing were performed using the ID-Micro Typing System gel test (Ortho Clinical Diagnostics), Anti-A and Anti-B reagents (Ortho Clinical Diagnostics), and A1 and B reagent RBCs (Ortho Clinical Diagnostics). Antibody identification was done using three different platforms: solid-phase (Immucor, Norcross, GA), automated gel (Grifols, Los Angeles, CA) and manual gel (Ortho Diagnostics, Raritan, NJ). Direct antiglobulin tests (DATs) were performed using an anti-IgG,-C3d polyspecific reagent and anti-IgG and anti-C3 mono-specific reagents (Immucor) through standard tube methods. DATs were performed with every new reaction pattern, first time identification of an alloantibody and/or NSP, and as needed with autoantibodies. Red cell elution studies were performed using the Gamma ELU-KIT II acid elution kit (Immucor, Norcross, GA). Elutions were performed with every positive DAT and when warm autoantibody was suspected.

Testing of Autoantibodies

Since autoantibodies can potentially cause nonspecific reactions, NSPs were assessed for cold and warm autoantibodies when such autoantibodies were suspected. Cold autoantibodies were identified by performing a “cold screen” by testing patient plasma with a three-cell antibody screen, autocontrol, and cord cells at room temperature. If a sample with NSP had a positive cold screen and a negative prewarm, routine tube or PEG screen, then the NSP result was removed except for actively pregnant or recently transfused (within 30 days) patients and was considered to be caused by cold autoantibodies.

Warm autoantibodies (WAA) were identified based on non-specific reactive patterns (positivity with all cells or some cells) with or without a positive autocontrol on a panel followed by a panagglutinin pattern seen in an elution, which was later verified by autoadsorption testing performed by a reference lab. Panreactive panels with panagglutinin patterns on an automated platform that was subsequently negative by tube or PEG testing were considered to be negative. If reactivity persisted on tube or PEG testing, samples were considered to have NSPs unless reference lab results confirmed the WAA with no other specific alloantibodies. However, for pregnant patients or recently transfused patients, NSPs were kept in their records regardless of reference lab testing results.

Statistical Analysis

The differences in reactivity strength across the three testing platforms were compared using the Kruskal-Wallis test. The number of cells used to establish a diagnosis of NSP was compared between two platforms using Mann-Whitney test. The distribution of ABO blood types between patients with NSP and all patients who underwent type and screen testing during the study period were analyzed using Fisher’s Exact Test. Statistical analyses were performed with GraphPad Prism v10.3.0, and p-values less than 0.05 were considered significant.

Results

Demographics & Study Population

During the 24-month study period, a total of 156,247 antibody screens were performed. Of these screens, 5,491 of these screens were positive, associated with the presence of antibodies in 1,992 unique patients. Of these patients, 1,583/1992 (79.5%) were female, yielding a male-to-female ratio of 1:3.87. Among the identified specificities, passive anti-D was the most frequent, detected in 840/5491 cases (21.6%), while anti-E was the most common alloantibody, detected in 477/5491 cases (12.2%). Nonspecific reactivity was identified in 257/1992 patients (13.0%) across 358/5491 panels (6.6%). Out of these, 87 NSP panels were identified using other manual methods such as tube testing or polyethylene glycol (PEG) testing. Consequently, 231 patients with 271/5491 (4.9%) antibody identification panels performed on gel and solid phase platforms were included for comparative analysis. The mean age of these patients at the time of first NSP detection was 54, with 170/231 patients (73.6%) being female (Table 1).

Table 1:

Demographics of Patients with NSP

	Immucor	Grifols	Ortho	Total
Count of Patients	93	115	51	231
Age at the time of first NSP (mean ± SD)	Not applicable	Not applicable	Not applicable	54 ± 22
Female gender (number, percent)	Not applicable	Not applicable	Not applicable	170 (73.6%)

Comparison of Platforms

Between 5/2021–4/2022, solid-phase was the primary automated system, and NSPs were identified in 95 screens out of 2692 positive screens (3.5%) in a total of 78,262 screens performed. Between 5/2022–4/2023, the facility migrated to an automated gel platform as the primary testing system, with NSPs identified in 123 out of 2799 positive screens (4.4%) among a total of 77,985 screens performed. During the 24-month period, whenever the pattern of reactivity observed in the original screen and panel could not be explained by the presence of specific alloantibodies, additional testing with a manual gel platform was performed concurrently. NSPs were also found in these manual gel screens in 24 out of the 95 solid-phase screens with NSPs and 29 out of the 123 automated gel screens with NSPs. The detailed comparison of these three platforms can be seen in Table 2.

Table 2:

Comparisons of One Solid Phase and Two Gel Based Platforms

	Immucor	Grifols	Ortho	Total
Count of Antibody Screens	78,262	77,985	Not applicable	156,247
Count of positive antibody screens	2692	2799	Not applicable	5491
Count of antibody identification panels with NSP identified on solid-phase or gel testing	95	123	53	271
Incidence of NSP among positive screens (percent)	95/2692 (3.5%)	123/2799 (4.4%)	Not applicable	271/5491 (4.9%)
Average number of cells screened for NSP (mean ± SD)	17.2 ± 7.1	14.4 ± 3.5	11.1 ± 4.8	14.7 ± 5.7
Total strength of reactivity for all panels with NSP (0–4) (mean ± SD)	0.75 ± 1.1	0.2 ± 0.5	0.5 ± 0.9	0.5 ± 0.9
Total strength of reactivity for all panels with NSP (0–4) Median (Q1-Q3)	0 (0–1)	0 (0–0.5)	0 (0–1)	0 (0–1)
Strength of reactivity for NSP cells (w-4) (mean ± SD)	1.7 ± 1.0	0.5 ± 0.4	1.4 ± 0.9	1.4 ± 0.9
Incidence NSP + Alloantibody (count, percent)	21/95 (21.6%)	21/123 (17.1%)	16/53 (29.8%)	58/271 (21.3%)
Incidence NSP + WAA (count, percent)	2/95 (2.1%)	9/123 (7.3%)	6/53 (11.3%)	17/271 (6.3%)
Incidence NSP + CAA (count, percent)	16/95 (16.8%)	40/123 (32.5%)	16/53 (30.2%)	72/271 (26.6%)
Incidence of NSP + Antibody (Allo, Auto or RHIg/MOAB) (count, percent)	40/95 (42.1%)	65/123 (52.8%)	34/53 (64.2%)	139/271 (51.3%)

NSP: Non-specific reactivity, WAA: Warm auto-antibody, CAA: Cold auto-antibody, RHIg: Rh₀(D) immunoglobulin, MOAB: Monoclonal antibody. SD: standard deviation.

Of tested panels, 139/271 (51.3%) had an additional reaction accompanying NSP. 72/271 (26%) were cold autoantibodies, 17/271 (6.3%) were warm autoantibodies, and 58/271 (21.4%) were one or more alloantibodies. The most common alloantibody to accompany NSP was anti-E in 14/271 panels (5.2%).

The mean strength of reactivity for NSP was 1.7 for the solid-phase platform, 0.5 for the automated gel, and 1.4 for the manual gel, and the differences in reactivity strength among the platforms were statistically significant (p<0.001). The average number of cells screened for NSP was 17.2 for the solid-phase platform, 14.4 for the automated gel, and 11.1 for the manual gel. Mann-Whitney test was done to compare solid-phase and automated gel and showed a significant difference among the platforms (p<0.01). Note that manual gel was not included in this analysis as it was a secondary platform.

In 30 testing events in which automated gel was positive for NSPs, only 11 manual gel panels tested positive, and in 16 testing events in which solid phase was positive, only 4 manual gel panels were positive. There was a single instance where NSP was detected in manual gel but not detected in solid phase. We also looked at patients who had testing done on all three platforms at different time points. Two patients had NSP on both solid phase and automated gel whereas manual gel was negative. One patient had NSP on solid phase and manual gel but tested negative on automated gel.

NSP and Alloantibody Formation

Nine panels that were performed on gel or solid-phase platforms (3.3%, 9/271) showed a correlation between the pattern of nonspecific reactivity and the later detection of a new specific alloantibody (Figure 1). New alloantibodies were identified as soon as the following day and as late as 7 months after the identification of NSPs, noting that the timing in these cases are affected by the number of intermediate screens performed. The alloantibodies detected included: anti-Jk^a in 3 patients (Patients 1, 3, 5), anti-D in 1 patient (Patient 2), anti-c in 1 patient (Patient 4), and anti-C + anti-e together in 1 patient (Patient 6). The mean and median strength of reactivity in these cases were 1.0. At the time NSP was identified, Patients 1, 3, and 5 had negative direct antiglobulin tests (DAT); Patients 2 and 4 did not have DAT results available. Patient 6 had a negative DAT on the automated gel but exhibited a weakly positive DAT with poly, IgG and C3 when tested on manual gel.

Figure 1:

Characteristics of non-specific reactivity in patients who subsequently developed specific alloantibodies.

Tube testing was utilized to address nonspecific reactivities on automated platforms. Some NSPs were resolved by PEG or routine tube testing shortly after the initial NSP detection (not shown in Figure 1). Of these, only three later developed into true antibodies (Figure 1): one anti-Jk^a on solid-phase ruled out by 3 PEG cells (Patient 1), one anti-Jk^a on automated gel ruled out by 3 PEG cells (Patient 5), and one anti-c on solid-phase ruled out by 11 PEG cells (Patient 4).

NSP and ABO Typing

The ABO typing of NSP patients were as follows: A: 63, B: 41, AB: 11, O: 115, Invalid:1 (previously A, with the ‘invalid’ typing reflecting recent ABO mismatched stem cell transplant patients). To determine if ABO typing of patients had any effect on the frequency of NSP seen, NSP patients were compared to all the patients who had a type/screen done during the study period, and no statistical difference was observed (p=0.12).

NSP and Pregnancy

There were 1773 patients who were either pregnant or in the postpartum period (3 months post-delivery) during the 24-month study period. Twenty-seven out of 1773 (1.5%) of these patients had NSP. Of these, 4/27 patients (14.8%) with NSP also had accompanying alloantibodies: anti-E (1 patient), anti-C (2 patients), and anti-K + anti-Kp^a (1 patient). Notably, one patient (Patient 3 in Figure 1) with a known anti-C exhibited NSP on the initial screen and later developed anti-Jk^a. There was no history of transfusion. The C and Jk^a antigen status of her baby was unknown; however, the cord blood DAT was negative and there was no evidence of hemolytic anemia.

Discussion

In this investigation, we examined rates of NSP by antibody testing platform at a single site where three popular testing modalities were employed in a large number of patients over a relatively short period. The most notable finding from the study was that solid-phase platforms exhibited stronger NSP reactivity compared to automated gel platforms, while gel platforms demonstrated a higher overall rate of nonspecific reactions. Manual gel testing, positioned as a secondary platform in our institution, displayed intermediate reactivity strength. However, the limited use of manual gel testing—due to its role as a secondary testing method—resulted in fewer panels and cells being tested, potentially affecting the comprehensiveness of the outcomes observed with this testing modality.

Another key observation was the low number of patients who subsequently developed an alloantibody, suggesting that NSPs most often likely represent a true ‘false positive’ rather than a developing (or evanescing) specific alloantibody just at the threshold of detection. It is important to note that our follow-up is only limited to patients who continued care at our institution and had subsequent screens performed. When NSPs preceded specific alloantibodies, the majority of those subsequently identified were targeting Rh antigens (such as C, c, e, D), findings similar to previous studies.^5,14 Notably, the presence of autoantibodies alongside NSP was seen quite frequently. In most instances, it was difficult to discern whether the NSP was attributable to these autoantibodies. Regardless, NSP cases were documented to ensure that wet crossmatching was carried out as a precautionary measure.

As previously reported, tube testing is generally considered less sensitive but more specific.^7,15,16 Three reactions that subsequently developed into true antibodies were positive on automated platforms but negative on tube testing. Two of these were anti-Jk^a, consistent with previous evidence that tube testing is less sensitive for identifying Kidd antibodies.² Since this study is not a validation analysis, it lacks reverse validation data, and thus conclusions regarding the sensitivity and specificity of tube versus automated methods cannot be drawn from our findings.

The study by Lu et al.¹⁷ examines NSP in pregnant patients and reports that 35% of NSP cases are associated with an alloantibody, a rate higher than observed in our population. Additionally, the antibodies identified in our pregnant patients—anti-C, anti-E, and anti-Jk^a—are known to be linked with hemolytic disease of the fetus and newborn. Notably, Patient 3, who experienced recurrent pregnancy losses, raises the question of whether alloimmunization might have contributed to these outcomes. In such cases, it is crucial to conduct follow-up ‘surveillance’ screens to ensure that NSP does not pose a clinical risk to either the mother or the fetus by developing into a specific alloantibody. Similarly, in Patient 6, identifying the anti-Jk^a earlier in pregnancy could have motivated closer fetal monitoring of hemolytic complications; fortunately in this case, the fetus did not seem to be affected. At our institution, any NSP detected in females of childbearing age are treated with caution, and crossmatch-compatible products are provided. For actively pregnant patients, regular interval follow-up screens are also advised as well as investigation as to whether NSPs alternatively reflect the presence of alloantibodies to ‘low incidence’ antigens.

Our institutional approach to resulting patients with NSPs involves several considerations, including the patient’s pregnancy status as well as history of transfusion and the need for transfusion. All pregnant patients and recently transfused patients keep their NSP regardless of what bench testing shows, which then requires full crossmatching. For all other patients, if NSP is not seen on a different testing modality such as manual gel or tube testing, the patient is recorded as negative and can be issued blood products by electronic crossmatch. A potential confounding factor to our analysis is that autoantibodies can cause reactivity patterns that are difficult to distinguish from a true NSP. However, our approach to both identifying warm and cold autoantibodies described in Methods minimizes this confounding. Patients with NSP-like reactivity from cold autoantibodies are considered negative when testing using prewarm screens are negative, and NSP-like reactivity due to warm autoantibodies are identified with autoadsorption studies. For patients who consistently show NSPs in certain platforms but negative reactivity in others, the preferred method that gives a negative result are recorded in laboratory information system and then used for future testing events.

In conclusion, nonspecific antibody reactions infrequently signal the development of a new alloantibody per this retrospective analysis. Among non-specific reactions, Immucor solid phase platform shows stronger reactivity by agglutination for NSPs, while Grifols automated gel shows a slightly higher rate of non-specific detection overall.