The Value of a Serum and Albumin Mixture for Use in the Detection of Blood Group Antigen-Antibody Reactions By

J Clin Pathol: first published as 10.1136/jcp.8.3.218 on 1 August 1955. Downloaded from

J. clin. Path. (1955), 8, 218.

THE VALUE OF A SERUM AND ALBUMIN MIXTURE FOR USE IN THE DETECTION OF BLOOD GROUP ANTIGEN-ANTIBODY REACTIONS BY

F. STRATTON AND E. R. DIMOND From the Regional Blood Transfusion Service, Manchester

(RECEIVED FOR PUBLICATION JANUARY 25, 1955)

It has been suggested by Wiener and Hurst was also included in the comparisons which were (1947) and Wiener, Hurst, and Sonn-Gordon made with respect to different blood group antigen- (1947) that mixtures of bovine albumin and human antibody reactions. plasma are more effective as diluents for the detec- Materials and Methods tion of incomplete antibodies than the widely employed 20% bovine albumin introduced by Red Cels.-Red cells were obtained in the form of clotted blood. The cells recovered from the clot were Diamond and Abelson (1945) or the human washed three times in normal saline and packed. The

plasma alone used by Wiener (1945). This has supernatant saline was completely removed from thecopyright. been confirmed by other workers, and Mollison packed cells. They were then resuspended at 10% (1951) has recommended mixing a suspension of strength in the appropriate diluting fluid. The test red cells in 20 % albumin with the test serum cells were of group 0 unless otherwise stated. diluted in group AB human plasma. Human Serum6-The serum used for diluting the The present investigation was undertaken to bovine albumin was obtained from donors of group determine whether the 20% bovine albumin test, AB and was either fresh or had been kept frozen solid for varying periods of time, and is referred to as ABhttp://jcp.bmj.com/ a which is open to criticism on number of grounds, serum. It was free from all blood group antibodies could be improved by substituting a mixture of and had no rouleau-forning properties. human serum and bovine albumin for the diluent. Antisera.-These human sera contained Rh or other Human plasma was not used because of clotting blood group antibodies of various strengths and had difficulties. After initial experiments it was decided been preserved frozen solid. to compare 20% bovine albumin with an equal Bovine Albumwi-Armour's 30% bovine albumin parts mixture of human serum and 30% bovine (or bovine plasma) was used in the tests. The 20% on September 26, 2021 by guest. Protected albumin. Saline, and often human serum alone, albumin was made by diluting this with half its

TABLE I RESULTS OF VARIOUS TESTS PERFORMED ON SERA CONTAINING WEAK ANTIBODIES (ANTI-D)

Tests Serum Agglutinins Titration of Serum in Titration of Serum in Papainizd 370 C. in 20% Albumin x Rlr Cells Mixture x No. Coombs Test ReaCllsRedCells Saline at 370 C. Serum-AlbuminCells at 370 C. RLr Rlr R2r rr (Rlr+R2r)* rr Rlr R2r rr 1:1 l: 2 1: 4 1: 8 Control 1:1 1: 2 1: 4 1: 8 Control 16 A14 ++ + + - ++...... --. - + - - 16 A43 + +I + + - - - - + w - 20 F 15 + + + + + + + - - w - w - - - - + - - 20 F 12 - + + + + - - - W-- - ++ + w i - 20G415 ++++ ++++ +++ - - - 20 G 44 +++ + + ++± +±*+- mixture. *Cell mixture. J Clin Pathol: first published as 10.1136/jcp.8.3.218 on 1 August 1955. Downloaded from

SERUM-ALBUMIN MIXTURE FOR DETECTING ANTlBODIES 219

volume of saline. The serum-albumin mixtures were Papainized Red Cells.-These were prepared and made by mixing the necessary volume of serum with used as described by Stratton (1953). 30% albumin. Technique.-The tests were performed in tubes Results 2 in. by l in. One volume, 0.05 ml., of antiserum Preliminary Tests.-A large number of pre- was placed in a tube and to this was added one volume were of a 10% suspension of packed cells in the appropri- liminary tests made to find the best mixture ate diluting fluid, i.e., 20% albumin, serum-albumin of human serum and 30% albumin to use as a mixture, saline, or AB serum. (These tests correspond diluent in the subsequent comparisons with 20% to the columns headed 1: 1 in the tables of results.) albumin. The choice of an equal parts by volume Titrations were performed by making serial dilutions, mixture was made for the following reasons. It viz., 1:2, 1:4, 1:8, and so on, of the antiserum in had the advantages of sensitivity and simplicity of the appropriate fluid and adding one volume of cells preparation. Albumin-rich mixtures were difficult suspended in the same fluid to one volume of each to work with and sedimentation of cells was slow dilution. All these tubes were then tapped to mix the in them owing to viscosity, while mixtures rich in contents and incubated normally at 37' C., in previ- serum lacked sensitivity. ously warmed racks for two hours. Controls were A series of experiments put up with each titration consisting of one volume relevant to disadvantages in the 20% albumin test of diluent and one of cell suspension in the diluent. will now be detailed, in which this equal parts Control titrations were also performed using Rh- mixture of serum and albumin was compared with negative cells in place of Rh-positive ones where the 20% albumin. The results are given largely in test was concerned with an anti-Rh serum. Similarly, tabular form and discussed in the final section. when other typing sera were being tested, positive and negative cells with respect to the particular antigen Detection of Weak Rh Antibodies.-Table I were used in parallel titrations as test and control. shows the results of testing sera containing weak The results were read under the low-power objective anti-D antibodies by various methods. It will be a of microscope after gently spreading the sedimented observed that two sera negative in the 20% copyright. clump from each tube on a slide with a pipette. albumin test were weakly positive in the serum- Results were scored as follows: albumin diluent, while two others giving very weak + + + + + =Massive agglutination with no free reactions in the former diluent gave stronger ones cells. in the latter. All the antibodies in this table were + + + + =Very strong agglutination with a few weak but were detected using papainized cells and free cells. by the Coombs test. The last two sera in the table

+ + + =Many large clumps of cells clearly contained antibodies not demonstrated in the http://jcp.bmj.com/ visible with more free cells. saline, 20% albumin or serum-albumin agglutina- + + =Multiple clumps (10-12 cells) just visible to the naked eye, with tion tests. many free cells. Enhancement of Titre of Rh-incomplete Anti- + =Small clumps (6-8 cells) visible micro- bodies by the Use of Serum-Albumin Mixture.- scopically; many free cells. This enhancement of titre has been described be- w = Clumps (3-4 cells) few in number; many free cells. fore and was found to be usually two- to eight- on September 26, 2021 by guest. Protected -= Negative; completely free cell sus- fold. Fourteen incomplete anti-D sera of various pension. titres were tested and all showed the greater effec- Anti-globulin or Coombs Test.-As described by tiveness of serum-albumin in this respect. The Mollison, Mourant, and Race (1952), with modifica- control tests using Rh-negative cells all gave nega- tions suggested by Stratton (1953). tive results. Table II illustrates a typical case.

TABLE II SERUM NO. 20 F 3 (INCOMPLETE ANTI-D) TITRATED AGAINST Rh-POSITIVE R1R1 CELLS IN VARIOUS DILUENTS AT 370 C.*

Dilutions DiluentDiluenti Control 111: 1 1:122 14!1: 4 1:8 1:16 1:32 1:64 1:128

20% albumin .. ++++ +++ ++ + - - - - Serum-albumin .. ..++++ ++++ +++ +++ ++ + ++ - Saline.-. . _ _ _ AB serum. +++ ++ + _

* Control titration against Rh-negative cells was negative throughout. J Clin Pathol: first published as 10.1136/jcp.8.3.218 on 1 August 1955. Downloaded from

_O1,, K

Ii CQJ CL _

-c' E LL1 U CQ i_ 0.-I_r . _0 r---I_+ -7 _ a Jo± ++~~~ z CQ0 a copyright. +~~~ L; 4- Su ij + + _ _-+--

_~*+I _-++ S - I http://jcp.bmj.com/

-J I + -4- 0: Ir1rlr 7LL YII I C .) I:)D C I I+ + LU 0 1+ ; -ra s . 4 + +4

LL. on September 26, 2021 by guest. Protected 7 + -+ I-v + LL

fC £E I~ E c I-I4- 1$ +X1 .2 M Hn z 0 CQ E.C E.i G

LLU

E U 'a+ =X , -"

| E u Eu ECa z-lI I K-v J Clin Pathol: first published as 10.1136/jcp.8.3.218 on 1 August 1955. Downloaded from

SERUM-ALBUMIN MIXTURE F OR DETECTING ANTIBODIES 221

Occurrence of Zoning in Detection of Rh Anti- lower titre was titrated with these high grade Du bodies.-Table III shows the titration of three cells, a negative result was obtained using 20% anti-Rh sera with various Rh-positive cells in bovine albumin, but a positive one using serum- different diluents. All these sera showed zoning albumin diluent (Table IV (d)). when the titration was performed using 20% Detection of Other Varieties of Rh Antibodies. albumin, but the zoning in each case was of a -Other varieties of Rh antibodies were found to different kind and degree. In Table III (a), the behave in a manner similar to the anti-D and zone using R2r test cells was a marked one in the anti-C+ D antibodies which have been considered middle of the titration; it did not occur so so far. An anti-c antibody (mainly incomplete) markedly using type RXR1 (CDe/CDe) cells since had a titre of 1 in 64 in 20% albumin, 1 in 256 the serum was of anti-C + D variety and zoning in serum-albumin, and 1 in 4 in saline against Rh- most commonly occurs only with respect to D. negative (cde / cde) cells, and was completely nega- The serum in Table III (b) showed zoning starting tive with RXR1 (CDe/CDe) cells in these media. earlier in the titration in 20% albumin. The third An anti-E antibody of titre 1 in 4 in albumin was serum (Table III (c)) shows only a slight zone of titre 1 in 16 in serum-albumin against R2R2 which is most marked in the lower dilutions, 1:1, (cDE/cDE) cells and control titrations were again 1:2, of the serum. It will be observed that the satisfactorily negative using R1R1 and Rh negative dilutions which gave negative results within the cells. zones of the 20 % albumin titrations gave un- Detection of Complete Antibodies.-Two varie- mistakable positive agglutination with serum- ties of complete antibodies were examined; first, albumin. In the case of the second and third what was considered to be a serum containing sera, if a single tube albumin test were used, the largely complete Rh antibodies, in view of the fact weak result at 1: 1 might readily have passed un- that its titre was the same in saline, 20% albumin, observed, whereas with serum-albumin the corre- and whole serum, namely, 1 in 32. The titre in sponding result was + + +. Serum 20 F 30 serum-albumin mixture was 64. copyright. (Table III (c)), being anti-D unlike the other two Anti-A and anti-B antibodies, some of which sera, showed a zone in 20% albumin against both had been selected as being of low titre in saline, R1 and R2 type cells. were titrated out in various diluents against cells Detection of Reaction between Incomplete Rh of appropriate group. There was no evident en- Antibodies and Du Cells.-It is known from pre- hancement of agglutination using either 20% vious work (Stratton and Renton, 1949) that Du albumin or serum-albumin mixture and, indeed, cells of low grade are not agglutinated by incom- with some sera the results were worse than when http://jcp.bmj.com/ plete Rh antibodies in a 20% albumin diluent. a saline diluent was used. Table IV (a) shows the titration of an incomplete Anti-Kell.-Six anti-Kell sera were titrated anti-D serum using various diluents and cells of with known Kell-positive cells in various diluents; type Rfir (CDue/cde) low grade. Here the ex- the results of a typical titration are shown in pected negative result was obtained using 20% Table V. It is interesting to observe the negative albumin, but with serum-albumin the result was positive. It is interesting to note that this titration TABLE V on September 26, 2021 by guest. Protected showed a zone. In Table IV (b) titration of the TITRATION OF AN ANTI-KELL SERUM (NO. 13 F I) WITH same serum with ordinary Rlr cells (CDe/cde) KELL-POSITIVE CELLS AT 37D C. IN VARIOUS DILUENTS using the same diluents is shown. This serum was specially selected and known to be valuable in Dilutions Diluent Control detecting Du positive cells when using the indirect 1:1 I 1:2 1:4 1:8 1:16 Coombs When a number of other sera technique. 20% albumin.. were used, some of lower titre, no agglutination Serum-albumin + + + w Saline was obtained even using serum-albumin diluent AB serum ..+++ + + +±+ +- and CDu/cde cells of low grade. Table IV (c) shows the results of titrating the same selected serum with cells of type R2r (cDuE /cde) high reaction in the 20 % albumin and the positive grade. Here an agglutination was obtained in result which was obtained in the serum-albumin both 20% albumin and serum-albumin diluents, titration; also an even stronger agglutination but the titre with the latter is much higher and the occurred in serum alone. Grove-Rasmussen individual results with each dilution stronger. (personal communication) informed us he had When a different, incomplete anti-D serum of found this using several different anti-Kell sera. J Clin Pathol: first published as 10.1136/jcp.8.3.218 on 1 August 1955. Downloaded from

222 F. STRATTON and E. R. DIMOND

Five out of six anti-KelU sera we examined showed Commentary serum than in any other this higher titre in whole The albumin test, which is widely used for diluent. routine purposes, suffers from certain defects. Anti-Duffy, Anti-S, and Anti-Kidd.-Several Some of these can be remedied by using serum- examples of each of these sera were tested with albumin mixture in the proportions suggested the various diluents, but all gave negative results instead of the 20% albumin mixture normally with cells containing the agglutinogen for these used. antisera. It has been shown (Stratton, 1953) that as many Anti-Lewis.-Various tests have been made as 5 to 7% of weak Rh antibodies in the maternal titrating anti-Lea and anti-Leb antibodies with the serum might be missed if the 20% bovine albumin appropriate cells at 160 C. and 370 C., using the test were used alone; when a serum-albumin mix- usual diluents. The results showed little enhance- ture is used many fewer Rh antibodies would ment of agglutination when 20% albumin or remain undetected. serum-albumin was used, compared with saline, All serological techniques for the detection of but it is probably advantageous to use an albumin blood group antigen-antibody reactions are liable technique at 370 C. if it is suspected that incom- to give false negative results due to the occurrence plete Lewis antibodies are present in a serum. of zoning, but the albumin test is particularly prone Effect of Albumin Mixture on Activity of Cold to this phenomenon. This means that when using Agglutinins.-Table VI shows the titration of a a 20% bovine albumin diluent a single tube test non-specific cold antibody at 80 C. and 370 C. cannot safely be employed and the test should be There is little difference between the titres and performed in the form of a titration using up to strengths of agglutination using 20% albumin and six tubes. The results reported here show that serum-albumin, but both give a higher titre than a serum-albumin diluent reduces the danger of these false negative results. Even where the first saline. copyright. tube of a serum titration shows a weak positive TABLE VI result, due to zoning when 20% bovine-albumin is used, the substitution of the serum-albumin TITRATION OF SERUM (NO. 12 D 17) CONTAINING A NON-SPECIFIC COLD AGGLUTININ WITH Rh-NEGATIVE mixture would result in a stronger positive result. CELLS AT 80 C. AND 370 C. IN VARIOUS DILUENTS It should not be assumed, however, that if a serum- albumin mixture is used every difficulty due to Dilutions Con- zoning will be eliminated in this test, but false Diluent _ trul http://jcp.bmj.com/ 1:1 1:2 1:4 1:8 1:16 1:32 negative results due to this cause will be consider- (a) 8' C.: ably diminished. 20% albumin ++++ +++ +++ ++ w _ Stratton and Renton (1949) showed that the Du Serum-albumin ++++ ++++++ ++ + _ Saline ..+++ ±++ + + type of cells, especially if of low grade, will not anti-D sera in bovine (b) 37° C.: react with incomplete 20% 20% albumin albumin, but if the Du type of cell is of high Serum-albumin Saline grade occasional positive reactions will be obtained depending upon the incomplete anti-D serum that on September 26, 2021 by guest. Protected is used. When a serum-albumin diluent is used, Centrifugation Experiments.-Experiments were depending upon the particular anti-D sera, some carried out to see whether time could be saved low grade Du positive cells may be detected; even during the performance of albumin tests by centri- so, a zone may occur in these titrations. Neverthe- fuging the tubes following a limited incubation less, it would seem that, in a population in which period and then reading the tests. An anti-D there occurs a large percentage of the Du-positive serum titrated in both 20% albumin and serum- type of person and where the albumin test is used albumin mixture was employed. The tubes were as a screening test in the first instance for the incubated for 20 minutes and then centrifuged at detection of D-positive persons, materially better low speed (500 r.p.m.) for five minutes and the results would be obtained if a serum-albumin results read. Comparison with the standard mixture were used for the original screening test. method showed that no serious error would result It should be emphasized that negative results would from using this method if a result is required need to be retested using the Coombs test. One quickly. Centrifugation experiments have been should not attempt to detect low-grade Du cells conducted with only a limited number of anti- using the albumin technique alone, whatever Rh sera and not with the full series of tests. diluent is used. J Clin Pathol: first published as 10.1136/jcp.8.3.218 on 1 August 1955. Downloaded from

SERUM-ALBUMIN AMIXTURE FOR DETECTING ANTIBODIES 223

Rarely, Rh antibodies are not detectable by the incubation time of 20 minutes at 370 C. in the antiglobulin or other tests, but only using the water-bath followed by low-speed centrifugation albumin technique. If these Rh antibodies are de- (500 r.p.m.) might well be used where speed is tected in the serum of an Rh-negative mother desired. Our tests with this technique have only during the antenatal period, the subsequent sample been concerned with a limited number of Rh usually gives a positive indirect Coombs test. antibodies. Anti-E antibody is one that is most often prefer- Many observers have shown that other sub- entially detected by the use of the albumin stances will act in the same way as bovine albumin technique. and human plasma or serum. Gelatin and Kell is an antigen to which persons readily be- dextran, for example, have been employed, and come immunized and anti-Kell antibodies are often one of us found, of many substances tried, that detected in human sera. The Coombs test has methyl cellulose was the best. All these substances most often been used for the detection of these suffered from the defect that they tend to cause antibodies, and the 20% bovine albumin test does false positive results which make the interpreta- not usually give a positive result. Our results tion of the test most difficult. The selection of a suggest, however, that if serum-albumin is used suitable test medium, therefore, involves finding positive results will be obtained. Grove-Rasmussen one which will not give false positive results but (personal communication) suggested that good re- will detect blood group antigen-antibody reactions sults were obtained if a serum diluent were used, with the minimum of difficulty. confirmed. and this has been In the authors' experience it is an advantage to Anti-Duffy, anti-Kidd, and sometimes anti-S add human serum in the proportions suggested to cannot be detected at all using the albumin tech- bovine albumin, and such a 50/50 mixture will not nique, whichever diluent is used, and the chances increase the false positive results or the difficulty of detecting Lewis antibodies are not materially of reading the test. The serum used in making copyright. increased by the substitution of a serum-albumin the mixture should be free from blood-group mixture for a 20% one. antibodies and all rouleau-forming properties, and In the case of anti-A and anti-B saline agglut- we are in favour of freezing it solid at least a week inins, our impression, after testing many samples, or ten days before making up the mixture. After was that these agglutinins are not as readily de- mixing it is stored at 40 C. and not kept for more tected in albumin or in serum-albumin mixtures as than a few days. http://jcp.bmj.com/ they are in saline, although they would be un- results in the first The reading and interpretation of the albumin likely to give completely negative cell and ones in saline. test have always presented difficulty. The two diluents positive removed from the tube Experiments were carried out to determine deposit should be carefully whether the increased rate of sedimentation of the and spread upon a microscope slide. At this stage cells was the sole reason for the enhanced aggregation of the red cells is liable to b2 present, red especially after centrifugation, and gives the effect agglutination of the serum-albumin mixture as on September 26, 2021 by guest. Protected compared with 20% albumin. These experiments of comets, i.e., a clump of cells from which a tail consisted of reading results of parallel titrations of cells can be seen trailing away. It is import- after incubation from one to four ant to leave the slide.a few minutes on the bench periods ranging of the micro- There was no evidence to before placing it under the low power hours. support this, scope and then to examine a moving film. This is in all cases the serum-albumin results were since done stage when the stronger than those where 20% albumin was used. by tilting the slide on the These experiments did suggest that economies in break-up of the cell aggregates will be observed, the test after one and a half the true agglutinates remaining intact. In the case time, such as reading often possible to hours' incubation, would not have such serious of saline agglutination tests, it is effects in preventing the detection of positive re- distinguish rouleau from true agglutination micro- sults if serum-albumin were used as a diluent com- scopically, but in the case of albumin or serum- pared with 20% bovine albumin. It seems to us, albumin tests it is quite impossible to make this therefore, that the time of incubation might be distinction. reduced accordingly when serum-albumin mixture The albumin technique is widely employed in is used. Low-speed centrifugation to reduce the the crossmatching test and often in the form of a time of incubation was also considered and an single tube test. While the authors would not J Clin Pathol: first published as 10.1136/jcp.8.3.218 on 1 August 1955. Downloaded from

224 F. STRATTON and E. R. DIMOND recommend this as a cross- routine method of REFERENCES matching, nevertheless it does seem a considerable Diamond, L. K., and Abelson, N. M. (1945). J. Lab. clili. lled., 30, advantage to use the serum-albumin mixture 204 and 668. suggested in place of 20% bovine albumin in such Grove-Rasmussen, M. Personal communication. Mollison, P. L (1951). Blood Transfusiont in Clintical .1ledicine. a test, and such a mixture would also be advan- Blackwell, Oxford. tageous if the test were done in the form of a Mourant, A. E., and Race, R. R. (1952). The Rh Blood Groups and Their Clinical Effects. Medical Research Council titration. In the crossmatching test, however, it Memorandum No. 27. H.M.S.O., London. seems preferable to use a saline agglutination Stratton, F. (1953). Lancet, 1, 1169. method in tubes at room temperature and at -and Renton, P. H. (1949). Brit. med. J., 2, 682. Wiener, A. S. (1945). J. Lab. clin. Med., 30, 662. 370 use C. coupled with the of the antiglobulin and Hurst, J. G. (1947). Exp. Med. Surg., 5, 285. technique. - and Sonn-Gordon, E. B. (1947). J. exp. Med., 86,'267. copyright. http://jcp.bmj.com/ on September 26, 2021 by guest. Protected