The Journal of International Medical Research 2011; 39:

Transcription

1 The Journal of International Medical Research 2011; 39: Comparative Evaluation of the Microcolumn Gel Card Test and the Conventional Tube Test for Measurement of Titres of Immunoglobulin G Antibodies to Blood Group A and Blood Group B D CHENG AND Y HAO Transfusion Department, First Hospital, China Medical University, Shenyang, Liaoning, China The conventional tube test (CTT) and microcolumn gel card test (MG) were used to measure immunoglobulin G (IgG) antibody titres to blood group A (anti-a) and blood group B (anti-b) in 288 serum samples. Diagnostic concordance and correlations between the two methods were analysed. MG was more sensitive than CTT in the detection of both anti-a and anti-b IgG. Correlation coefficients for the two methods for the titration of anti-a and anti-b were and , respectively. The regression equation representing the relationship between the two methods was calculated and used to convert critical values for anti-a and anti- B IgG antibodies determined by CTT to critical values for MG titration. Analyses showed 86.8% and 89.9% diagnostic concordant rates for the detection of anti- A and anti-b IgG, respectively. It was concluded that titration results obtained by MG correlate well with those obtained by CTT, but critical values determined by CTT need to be converted into corresponding MG values. KEY WORDS: MICROCOLUMN GEL CARD TEST; CONVENTIONAL TUBE TEST; IMMUNOGLOBULIN G; ANTIBODY TO BLOOD GROUP A; ANTIBODY TO BLOOD GROUP B; BLOOD GROUP Introduction The titre of an antibody represents the strength of that antibody s response to a specific antigen. Titres are generally determined with a semi-quantitative assay by serial double-fold dilution. 1 Several different techniques have been employed for titration, of which the conventional tube test (CTT) and the microcolumn gel card test (MG) are the most frequently used. Historically, CTT has been used as the standard technique for immuno - haematological studies, such as in the direct antihuman globulin test for the diagnosis of autoimmune haemolytic anaemia and also in screening for specific antibodies in transfusion medicine. 2 7 Based on CTT, a number of clinically relevant cut-off values have been set. 8,9 CTT is time consuming, involves many cell washing steps and 934

2 requires relatively experienced personnel to interpret the results. It is also difficult to automate and standardize and, thus, its use has been challenged by the newer MG technique. 10 MG is an easy and sensitive technique that requires no red cell washing and uses gel filtration media impregnated with an antihuman globulin reagent to induce agglutination Due to the simplicity, reproducibility and sensitivity of MG, many blood banks and transfusion departments are now gradually adopting this technique. 14,15 Although titration has long been considered to be one of the poorest techniques with respect to standardization and automation, as well as reproducibility, it is still necessary in several medical fields. In transfusion medicine, titration is routinely used to monitor blood group antibody titres in pregnant women to prevent newborn haemolytic disease and to evaluate levels of incompatible blood group antibodies before transfusion of pooled platelet concentrates from incompatible donors are given to recipients. In transplantation programmes, titration is used to evaluate the effectiveness of pretreatment regimens designed to reduce antibody levels before incompatible allogeneic transplantation and to monitor relative antibody levels over time to prevent the occurrence of rebound. 3,16 18 Antibody titration levels are assessed relative to appropriate critical values. Since a number of antibody titre critical values were originally established using CTT, it needs to be determined whether these values can be applied to antibody titres measured using MG techniques. In the present study, the titres of immunoglobulin G (IgG) antibodies against blood group A (anti-a) and blood group B (anti-b) in frozen serum samples from pregnant women were measured concurrently using CTT and MG in order to compare the sensitivity of each technique. The correlation and diagnostic concordance of the results from the two methods were also compared. Subjects and methods PREPARATION OF SERUM SAMPLES Serum samples containing anti-a IgG or anti-b IgG from blood type O and Rhesus D- positive pregnant women whose partners were also Rhesus D-positive and had blood types A, B or AB were collected and stored at 30 C. Written informed consent was obtained from the study participants and the study protocol was approved by the China Medical University Ethical Board. Serum samples and commercial reagent type A, B and O red blood cells (Shanghai Hemo-Pharmaceutical and Biological Co., Shanghai, China) were prepared for the titration tests as described previously. 2 Briefly, reagent red blood cells were diluted to 0.8% and 4% suspensions for MG and CTT, respectively. Frozen serum samples were thawed immediately before testing and then 400 µl of serum was mixed with 400 µl 0.2 mol/l 2-methyl-diethyltryptamine (2-ME) solution (Changchun Brother Biotech Co., Changchun, China) and incubated at 37 C for 1 h to cleave IgM blood group antibodies. Type O plasma was used as a control and completeness of the cleavage was assessed by microscopy after the addition of A or B blood cell reagents. The treated serum samples were serially two-fold diluted with physiological saline from 1 : 2 to 1 : CTT TITRATION For the CTT, 100 µl of each of the serially diluted solutions of 2-ME-treated serum samples and 50 µl of 4% reagent red blood cells were added to marked tubes, mixed and then incubated at 37 C for 30 min. The samples were then centrifuged at 840 g for 935

3 10 s and washed three times with 0.9% saline. The supernatants were completely decanted, one drop of antihuman globulin reagent (Immucor, Norcross, GA, USA) was added to each tube and the samples were then centrifuged again at 840 g for 10 s. The results were read macroscopically using a weak positive result, defined by the American Association of Blood Banks as tiny aggregates with a turbid reddish background, 19 as the endpoint; the titre value was the reciprocal of the highest dilution that gave a weak positive result. MG TITRATION A semi-automatic blood processing system with DG Gel Coombs cards containing antihuman globulin (Diagnostic Grifols, Barcelona, Spain) was used for the MG titration. Briefly, 25 µl 0.8% reagent red blood cell suspension and 50 µl of the serially diluted solutions of 2-ME-treated serum samples were added to each microcolumn. The mixture was incubated at 37 C for 15 min and then centrifuged at 900 g for 10 min. A 1+ reaction, defined by the manufacturer as the presence of some small sized agglutinations in the column, was used as the endpoint; the titre value was the reciprocal of the highest dilution that gave a 1+ result. STATISTICAL ANALYSES The number and percentage of subjects with each titre was recorded. Log-converted mean values of anti-a and anti-b IgG for the MG and CTT methods were analysed using a paired t-test. The non-parametric Spearman test was used to evaluate the correlation between the results from the two methods and linear regression analysis was performed using the Passing Bablok method. The regression equation was calculated and estimates of the slope and intercept expressed with 95% confidence intervals (CIs). Using the regression equation, the critical value for the MG assay that corresponded to a critical value of 64 for the CTT assay was calculated and the diagnostic concordance of the two methods was analysed using the κ test. A P-value < 0.05 was considered to be statistically significant, and a κ > 0.75 was considered to be perfect concordance. Statistical analyses were performed using SPSS version 15.0 (SPSS Inc., Chicago, IL, USA) and MedCalc version 3.0 (MedCalc, Mariakerke, Belgium) statistical software. Results A total of 288 different serum samples were tested. Anti-A and anti-b IgG titres measured by CTT and MG are shown in Tables 1 and 2, respectively. The titres of both anti-a and anti-b IgG determined by MG were frequently approximately two-fold higher (one more dilution step) than those determined by CTT, and several MG results were more than two-fold higher than the corresponding CTT results. This is illustrated visually with the percentage distributions of the anti-a and anti-b IgG titres measured by the two methods shown in Fig. 1. The mean anti-a IgG titre was 2.17 times higher with MG than with CTT (187 and 86, respectively); this difference was statistically significant (P < 0.001). The mean anti-b IgG titre was 2.61 times higher with MG than with CTT (137 and 52, respectively); this difference was also statistically significant (P < 0.001). LINEAR REGRESSION ANALYSIS As shown in Fig. 2A, the anti-a IgG titre results determined by MG and CTT were significantly correlated, with a correlation coefficient of (P < 0.001). Passing Bablok regression analysis showed a linear 936

4 TABLE 1: Numbers of each titre of immunoglobulin G antibody to blood group A measured by the conventional tube test (CTT) and the microcolumn gel card test (MG) in 288 individual serum samples CTT titre MG titre

5 TABLE 2: Numbers of each titre of immunoglobulin G antibody to blood group B measured by the conventional tube test (CTT) and the microcolumn gel card test (MG) in 288 individual serum samples CTT titre MG titre

6 A MG method CTT method 20 Titres (%) B Anti-A titres MG method CTT method Titres (%) Anti-B titres FIGURE 1: Percentage distribution of titres of immunoglobulin G antibody to (A) blood group A (anti-a) and (B) blood group B, measured by the conventional tube test (CTT) and the microcolumn gel card test (MG) in 288 individual serum samples relationship between the two methods that can be represented by the equation log(y 1 ) = log(x 1 ). The y-intercept and slope, with associated 95% CIs, were (95% CI , ) and (95% CI , ), respectively. The anti-b IgG titre results determined by MG and CTT were also significantly correlated (Fig. 2B), with a correlation coefficient of (P < 0.001). The linear regression equation was log(y 1 ) = log(x 1 ). The y-intercept and slope, with associated 95% CIs, were (95% CI , ) and (95% CI , ), respectively. DIAGNOSTIC CONCORDANCE When the critical value for anti-a and anti- B IgG titres determined by CTT was set at 64, the critical values for anti-a and anti-b IgG titres determined by MG calculated from the regression equation were 136 and 142, 939

7 A MG titres anti-a [log(y 1 )] B 3.5 CTT titres anti-a [log(x 1 )] 3.0 MG titres anti-b [log(y 1 )] CTT titres anti-b [log(x 1 )] FIGURE 2: Scattergrams and linear regression analysis of log-converted titres of immunoglobulin G antibody to (A) blood group A (anti-a; r = ; P < 0.001; regression equation log(y -1 ) = log(x -1 )) and (B) blood group B (anti- B; r = , P < 0.001; regression equation log(y -1 ) = log(x -1 )), measured by the conventional tube test (CTT) and the microcolumn gel card test (MG) (the solid lines correspond to the regression equation line of best fit and the dotted lines show the 95% confidence interval) respectively. In terms of the titres available, a critical value of 128 was, therefore, used for both the anti-a and anti-b IgG titres, with titres of 128 or higher indicating a positive result. Using these critical values, the diagnostic concordance rate (positive or 940

8 negative results with both tests) for the anti- A IgG titres was ([ ]/288) 100 = 86.8% (Table 3), with κ = (P < 0.001). The diagnostic concordance rate for the anti- B IgG titres was ([ ]/288) 100 = 89.9% (Table 4), with κ = (P < 0.001). These κ values indicate a level of diagnostic concordance in the perfect range for both antibodies. Discussion Antibody titres reflect the reactive ability against a specific antigen. Traditionally, titration has generally been approached semi-quantitatively using serial two-fold dilution assays. 1 ABO blood group antibody titre monitoring is used in prenatal testing 3,16 18 and transplantation programmes. TABLE 3: Diagnostic concordance of titres of immunoglobulin G antibody to blood group A measured by the conventional tube test (CTT) and the microcolumn gel card test (MG) using critical titrations of 64 for CTT and 128 for MG CTT MG Positive Negative Total Positive Negative Total TABLE 4: Diagnostic concordance of titres of immunoglobulin G antibody to blood group B measured by the conventional tube test (CTT) and the microcolumn gel card test (MG) using critical titrations of 64 for CTT and 128 for MG CTT MG Positive Negative Total Positive Negative Total Techniques for antibody titration have undergone extensive development and several methods are routinely employed for ABO blood group antibody titration, 1,2 with CTT and MG being the most frequently used. CTT is considered the classical test for antibody titration but is time consuming, because of the extensive cell washing required, and is difficult to standardize. 10 Owing to the simplicity, reproducibility and sensitivity of the MG method, many blood banks and transfusion departments are now gradually adopting this technique. 14,15 It is important that results obtained from different laboratories should be transferable and comparable, and that results are interpreted in the same way, to enable accurate clinical diagnosis and intervention. For clinical situations, a critical value is often used as a reference point. Such critical values are specific to the technique used and cannot always be applied to other methods. It has been reported that MG is more sensitive in detecting serum antibodies, 20 resulting in higher titres for the same batch of samples relative to CTT results. 9 Studies of these two methods have mainly focused on the sensitivity and specificity of the techniques, with little attention being paid to the appropriateness of interpreting MG results in relation to critical values determined using CTT. The relative sensitivity of the two methods demonstrated in the present study is consistent with that in previous reports. 4 The sensitivity of MG was greater than that of CTT for the titration of both anti-a and anti- B IgG antibodies, with the MG results being approximately two-fold higher (one more dilution step) than those for CTT. Furthermore, the MG titration results were primarily clustered around values that were two-fold higher than the CTT titration results for the same samples. None of the MG titres was lower than the corresponding CTT titre. 941

9 This is in contrast to the results of a study comparing MG and CTT for anti-d titration, 9 in which MG and CTT results for the same samples showed a randomized distribution. It is possible that this difference was related to the use of a 1+ endpoint for CTT; recent studies have shown that the use of a weak positive endpoint for CTT, as used in the present study, may decrease the variance in titre determination. 2,21 In addition, in the present study, one experienced technologist was responsible for all of the testing and interpretation of the results; this may have contributed to a lower variance and resulted in a clearer distribution of MG results compared with CTT results. This low variance also allowed the correlation between the two methods to be analysed further. In the present study there was a strong correlation between the MG and CTT results for both anti-a and anti-b IgG titres. Linear regression analysis, however, suggested that the same critical values used with the CTT method are not appropriate for use with the MG method. Using CTT, an anti-a or anti-b IgG titre of 64 is generally considered in China to be the critical value for clinical intervention in pregnant women to prevent ABO blood group-related neonatal haemolytic diseases. When the critical value for anti-a and anti-b IgG titres determined by CTT was set as 64, the critical values for the anti-a and anti-b IgG titres determined by MG were estimated to be 136 and 142, respectively, using the regression equations established in the present study. As the MG titration method becomes more widely used in prenatal testing, transplantation programmes and transfusion medicine, it is important that the results are not interpreted using critical values determined by CTT. To the best of our knowledge, very few studies have been conducted to determine whether the critical values of anti-a and anti-b IgG antibodies based on CTT are applicable to MG techniques. The present study found a good correlation between CTT and MG results for both anti-a and anti-b IgG titres. The critical values for anti-a and anti-b IgG antibodies determined by MG were established by converting the critical values for CTT titration using the regression equation. Using the calculated critical values, the results of anti-a and anti-b IgG titres determined by MG showed perfect diagnostic concordance with those determined by CTT, showing that MG can be used for titration assays as long as the critical values determined by CTT are converted into the corresponding MG values. Data from the present study will be useful in defining the critical values for the MG technique for the titration of anti-a and anti- B IgG antibodies. Further investigations should focus on analysis of the relationship between CTT and MG for the titration of other kinds of antibodies. In addition, the participation of multiple investigation centres would enable universal guidelines to be established, reducing variance between laboratories and ensuring that results obtained by different laboratories are transferable and comparable. Acknowledgement This study was supported by a grant from the 2008 Fund Project of the Department of Education, Liaoning Province, China (Topic No ). Conflicts of interest The authors had no conflicts of interest to declare in relation to this article. Received for publication 25 January 2011 Accepted subject to revision 1 February 2011 Revised accepted 7 March 2011 Copyright 2011 Field House Publishing LLP 942

10 References 1 AuBuchon JP, de Wildt-Eggen J, Dumont LJ: Reducing the variation in performance of antibody titrations. Arch Pathol Lab Med 2008; 132: Shirey RS, Cai W, Montgomery RA, et al: Streamlining ABO antibody titrations for monitoring ABO-incompatible kidney transplants. Transfusion 2010; 50: Cooling LL, Downs TA, Butch SH, et al: Anti-A and anti-b titers in pooled group O platelets are comparable to apheresis platelets. Transfusion 2008; 48: Coombs RR, Mourant AE, Race RR: A new test for the detection of weak and incomplete Rh agglutinins. Br J Exp Pathol 1945; 26: Coombs RR: Historical note: past, present and future of the antiglobulin test. Vox Sang 1998; 74: Das SS, Chaudhary R, Khetan D: A comparison of conventional tube test and gel technique in evaluation of direct antiglobulin test. Hematology 2007; 12: Lange J, Selleng K, Heddle NM, et al: Coombs crossmatch after negative antibody screening a retrospective observational study comparing the tube test and the microcolumn technology. Vox Sang 2010; 98: e269 e Kamesaki T, Oyamada T, Omine M, et al: Cutoff value of red-blood-cell-bound IgG for the diagnosis of Coombs-negative autoimmune hemolytic anemia. Am J Hematol 2009; 84: Novaretti MC, Jens E, Pagliarini T, et al: Comparison of conventional tube test with diamed gel microcolumn assay for anti-d titration. Clin Lab Haematol 2003; 25: Reis KJ, Chachowski R, Cupido A, et al: Column agglutination technology: the antiglobulin test. Transfusion 1993; 33: Nathalang O, Chuansumrit A, Prayoonwiwat W, et al: Comparison between the conventional tube technique and the gel technique in direct antiglobulin tests. Vox Sang 1997; 72: Lapierre Y, Rigal D, Adam J, et al: The gel test: a new way to detect red cell antigen antibody reactions. Transfusion 1990; 30: Bromilow IM: Gel techniques in blood group serology. Med Lab Sci 1992; 49: Dittmar K, Procter JL, Cipolone K, et al: Comparison of DATs using traditional tube agglutination to gel column and affinity column procedures. Transfusion 2001; 41: Grey D, Connolly M, Erber WN: Comparison of low ionic diluents for use with the Diamed antiglobulin gel test. Transfus Med 2002; 12: Moise KJ Jr: Management of rhesus alloimmunization in pregnancy. Obstet Gynecol 2002; 100: Tydén G, Donauer J, Wadström J, et al: Implementation of a protocol for ABOincompatible kidney transplantation a threecenter experience with 60 consecutive transplantations. Transplantation 2007; 83: Tobian AA, Shirey RS, Montgomery RA, et al: The critical role of plasmapheresis in ABOincompatible renal transplantation. Transfusion 2008; 48: American Association of Blood Banks (AABB): Technical Manual, 13th edn. Bethesda: AABB, Paz N, Itzhaky D, Ellis MH: The sensitivity, specificity, and clinical relevance of gel versus tube DATs in the clinical immunology laboratory. Immunohematology 2004; 20: AuBuchon JP, de Wildt-Eggen J, Dumont LJ: Reducing the variation in performance of antibody titrations. Vox Sang 2008; 95: Author s address for correspondence Dr Daye Cheng Transfusion Department, First Hospital, China Medical University, 155 North Nanjing Street, Shenyang , Liaoning, China. dayecheng_cmu@yahoo.cn 943