- Email: [email protected]
Inhibition of Rh antibodies
Inhibition of Rh antibodies WARREN H. PEARSE, M.D. CALVIN HOBEL, M.D. Omaha, Nebraska
S I N c E the etiology of the long puzzling condition, erythroblastosis fetalis, was discovered in 1940, many attempts at modification or elimination of the disease have been made. These treatment attempts basically fall into four categories: ( 1) Initial sensitization of the mother may be prevented by protecting placental integrityl 2· 13 or avoiding intrauterine trauma at delivery 6 ; (2) presumed or demonstrated sensitization may be blocked by injection of antibody to neutralize fetal cells which have gained entrance to the maternal circulation 5 ; ( 3) attempts may be made to neutralize or inactivate demonstrated circulating antibody in the maternal circulation before significant amounts cross the placenta (previous efforts in this area include the use of Rh hapten, 3 cortisoP 1 and ethylene disulfonate) 15 and (4) the affected infant is accepted, perhaps delivered preterm, and, if necessary, exchange transfusion is carried
fate, adenylic acid, and uridylic acid in 2 per cent concentration at pH 6.8 to be inhibitors of an ti-D (Rho) and this was later extended to anti-c (hr') and anti-e (hr") .9 A large series of other compounds tested, including sugars, amino acids, desoxyribonucleic acid (DNA) derivatives, purines, and pyrimidines showed no similar effect. The report of Boyd, McMaster, and Wasczenko-Zacharczenko2 demonstrated inhibition of anti-D by high concentrations (0.1 to 0.2 molar) of "unnatural sugars," especially !-glucose, 1mannose, and d-gulose. Since the saline antibodies, whose inhibition was demonstrated by these authors, are not transplacentally transmitted, it seemed desirable to extend these studies to the albumin reacting or "incomplete" antibodies ( 7S) which are the pathogenic agents in erythroblastosis fetalis. Further, while such in vitro studies give valuable evidence concerning the chemical structure of red cell surface antigens, and antigen-antibody reactions, it also seemed desirable to extend the investigation to in vivo studies, with the ultimate aim of aiding the problem of Rh isoimmunization in the human.
---.&- 4 UUL.
The present study considers a new application of the third approach. Inhibition of antiRh antibodies (as well as anti-Lutheran) by ribonucleic acid derivatives was first reported by Hackel, Smolker, and Fenske. 7 These studies, carried out entirely in saline media, demonstrated cytidylic acid, cytidine hemisul-
Materials and methods
Inhibition titers. Compounds to be tested for antibody inhibiting effect were prepared in varying concentrations from 1 to 20 per cent in 0.2M phosphate buffer. Tenth normal potassium hydroxide was used to adjust pH to either 6.8 or 7.4. Standard anti-D sera of the saline reacting and albumin reacting types as well as patients' sera were serially diluted with normal
From the Department of Obstetrics and Gynecology, University of Nebraska College of Medicine. Supported in part by Grant No. 2G-631 from the United States Public Health Service. Presented at the Thirty-first Annual · Meeting of the Central Association of Obstetricians and Gynecologists, Denver, Colorado, Sept. 12-14, 1963.
46
Volume 39
Inhibition of Rh antibodies
.!\umber 1
Table I. Example of the Race agglutination scale
Table II. Sera inhibited with ribonucleic acid derivatives ( 8 to 20 per cent) in 11 patients
Agglu- 1: 1 1:2 1:4 1:8 1: 16 1:32 1:64 1: 128 tination 10 10 10 8 5 2 0 0 Saline anti-D serum + 2 per cent suspension 0+ red blood cells in saline Total score
= 45
Average
Saline Albumin Coombs
Control Race score
Treated Race score
0 24
0
44
22
Control A and B serum
saline or type AB serum through ten tubes, the volume in each tube being 0.1 ml. Onetenth milliliter of the test solution was added to each tube and the mixture incubated at 37° C. for 30 minutes. One-tenth milliliter of type 0 Rh-positive red blood cells in 2 per cent suspension in either saline or protein media as appropriate, was added to each tube and incubated at 37° C. for 60 minutes. Positive controls utilizing saline instead of the experimental compound and negative controls utilizing type 0 Rhnegative red blood cell suspensions were routinely employed. After 60 minutes, the tubes were centrifuged at 1,000 r.p.m. for 1 minute, the cells resuspended by gentle tapping and the tube contents examined microscopically. Agglutination was graded by the Race scale/ 6 ranging from a score of 10 for agglutination clearly visible to the naked eye, to 8 for large microscopic agglutinates, 5 for average size agglutinates, 3 for small agglutinates, 2 for the smallest definite agglutinates seen microscopically, and 0 for no evidence of agglutination (Table I). In most instances, two examiners graded the slides independently. When no agglutination was demonstrated in the saline system, the cells were washed three times in normal saline with low speed centrifugation, 0.1 ml. of Coombs antihuman globulin was added, the tube was centrifuged at 500 r.p.m. for 1 minute and re-examined microscopically as noted above. The same donor was used as a source of type 0 Rh-positive cells throughout the nucleic acid derivative tube and animal studies. Varying donors were employed for the remainder of the study.
47
2
= no inhibition
Table III. Compounds* inhibiting albumin antibodies in the serum of Patient M. S. Race score Saline
Albu-1 min Coombs
Control 0 38 62 Glucose 10 per cent 20 60 Citric acid 10 per cent 10 66 Succinic acid 10 per cent 0 58 50 Reticulose 8 60 0 Glycine 10 per cent Ethylene diacetate 10 -hemolyzed cellsper cent *Tests of these compounds against saline Anti-D test sera demonstrated no inhibitory effect.
Erythrocyte antigen for rabbit injection. Rh-positive red cell antigens for rabbit injection were prepared in two fashions: 1. Belkin and Weiner1 demonstrated that suspensions of cell stromata retained their antigenic properties. Type 0, Rh genotype Dee/deE red blood cells were washed in isotonic saline until a clear supernatant was obtained. The cells were then lysed in distilled water with three washings, the ghost cells centrifuged and mixed with normal saline to provide a 1/10 dilution. 2. Lubinski and Portnuff14 pointed out that heating Rh-positive cells suspended in saline to 56° C. rendered the cells Rh negative. Hubinot10 then demonstrated that an Rh substance was present in the supernatant saline by inhibition studies. The same type 0, Rh genotype Dee/deE red blood cells were washed in isotonic saline until clear, and with an equal amount of saline added the mixture was adjusted to pH 7.2 by adding
48
Pearse and Hobel
..\.111.
INHIBITION OF SALINE ANTI-D TEST SERA
§ ... g
15
v
;::
~
Results
25 . , ,
20
\
' • \\
10
·'\.-..
~CONTROL
May 1, IY64 Obst. & GYnec.
Control rabbits received the buffer solution alone as a subcutaneous injection over a '1 ""'! , • .l snnuar 1 uay penuu.
~=l ~
J.
10
2
DRUG CONCENTRATION
0
12
IPIR
14
16
18
20
CENT)
= ADENYLIC ACID
· = CYTIDINE HEMISULFATE Fig. 1.
O.lN KOH. The cells were then heated for 17 minutes at 56° C., centrifuged, and the supernatant fluid recovered. This was mixed with normal saline to provide a 1/10 dilution. Animal studies. Control studies of naturally occurring agglutinins to 0 Rh-positive human red cells in rabbit serum were carried out. Six rabbits were then injected subcutaneously with i mi. of cell stromata preparation on days 1, 2, 3, and 8, 9 and 10. Six rabbits were sensitized in the same fashion with the heated supernatant preparation. Three control rabbits were injected in the same fashion with a cell stromata preparation of type 0, Rh-negative cells. A second titer level was obtained on day 7, a third on day 17, and a fourth on day 29. Specificity of the induced antibodies in the rabbit's sera was evaluated by selective absorption and tests against a panel of known human cells. As noted below, specificity could not be demonstrated. Antibody inhibition or inactivation was attempted by injecting subcutaneously 1.0 ml. of a test compound daily on days 32 to 39 and again determining antibody levels on day 40.
1. The four RNA derivatives originally proposed (adenylic acid, uridylic acid, cytidylic acid and cytidine hemisulfate) were tested against commercial an ti-D (Rho) serum and against individual patient sera. Extensive initial testing in saline revealed minimal effect in 2 per cent concentration and this was evident only in serum dilutions. Fig. 1 illustrates results characteristic of the four compounds. The adenylic acid 5' did not maintain effectiveness more than 24 hours unless frozen, whereas the other compounds retained activity as long as 6 months at 4° C. In protein media, and utilizing appropriate cell suspensions in AB serum all four cornpounds were equally effective in inhibiting "albumin" antibodies. However, the "incomplete" antibodies demonstrated by Coombs antihuman globulin persisted, although reduced, when saline ("complete") antibodies were absent. Results on 11 serum specimens from patients vvith kno\vn erythroblastotic infants are outlined in Table II. Saline titers in sera obtained from sensitized patients wPn' generally of low magnitude ( 1 : 1 or 1 : 2) and usually disappeared even though the sera were frozen. Compounds were also tested against anti-A, anti-B, anti-M, and anti-?\ test sera. There was no inhibiting effect on agglutination in any instance. 2. A search was begun for more effective inhibiting compounds. Among those tested were glucuronic acid ( 10 per cent), glucose ( 10 and 20 per cent) , sucrose ( 10 and 20 per cent), citric acid ( 10 per cent) , succinic acid ( 10 per cent), reticulose (a "nucleic acid complex"), and glycine ( 10 per cent i. Surprisingly, while none of these compounds were effective against saline anti-D test sera, nor in inhibiting Coombs antibodies in patient serum, they all seemed to have an inhibitory effect in albumin media (Table III!. Ethylene diacetate was tested since a similar compound, ethylene disulfonate, had been suggested as an antibody inhibitor in minute
Volume 89 Number 1
Inhibition of Rh antibodies
49
Table IV. Test sera inhibited with sugar amine derivatives* 10 per cent
I
Race score control
10 per cent glucosamine
n-acetyl glucosamine
galactosamine
10 per cent n-acetyl galactosamine
33 17
0 0 0
17 2
0
20
0
2
Saline Albumin Coombs
10 per cent
0
*Control A and B serum = no inhibition. Equivalent results at pH 6.8 and 7.4.
Table V. Rabbit saline titers-Race score
Titer
Rabbits 1 through 6
I (control)
10 25 20
6 16 36 28
4 8 21 19
19 25 10
26 42 13
12
1-
II (day 7) III (day 17)
IV (day 29) v (day 40) DG* CH* CON*
Rabbits Rabbits 7 through 13 through 12 15
=
4
13
D-glucosamine, 10 per cent; CH == cytidine *DG buffer control. hemisulfate, 15 per cent; and CON
=
Table VI. Rabbit albumin titers-Race score
Titer
Rabbits 1 through 6
I (control)
II (day 7) III (day 17
IV (day 29) V (day 40) DG* CH* CON*
I7Rabbits through 12
Rabbits 13 through 15
6
9
4
11 23 20
19 36 27
14 21 18
15 18 18
11 18
6 4
3
14
*DG = D-glucosamine, 10 per cent; CH = cytidine buffer control. hemisulfate, 15 per cent; and CON
=
Table VII. Ten per cent glucosamine in patient
quantitiesY In a 10 per cent solution, all red cells were hemolyzed. A series of compounds was discovered, however, which appeared to inhibit saline, albumin, and Coombs antibodies more effectively than the ribonucleic acid derivatives. These were glucosamine, N-acetyl glucosamine, galactosamine, n-acetyl galactosamine, and pyridoxal phosphate. The acetyl derivatives were found less effective and not extensively tested (Table IV). Further multiple studies with anti-D and anti-C commercial sera and patient sera confirmed the effectiveness of glucosamine and galactosamine in all three antibody systems. Animal studies. Animal results were disappointing. Naturally occurring antibodies to human 0 Rh positive cells were present in an average Race score of 5 in saline and 7 in protein media. Rabbits were readily immunized to the cell stromata preparation and supernatant fluid preparation, but specificity of the antibodies could not be demonstrated with the use of panel cells or by selective absorption. Immunization with the cell stromata preparation produced equivalent sensitization utilizing both Rh-positive and Rh-negative cells. The heated supernatant preparation produced higher antibody titers, but only Rh-
J- M.
Control serum Rh - cells
Rh + cells (titer)
Glucosamine tube Rh + cells (titer)
Posttreatment sera Rh + cells (titer)
0
33 ( 1 :4) 75 (1:64) 92 (1:256)
10 (0) 18 ( 1: 1) 32 (1:2)
32 ( 1: 4) 68 (1:32) 85 (1:128)
Saline Race score Albumin
0
Coo1nbs Race score
0
50 Pearse and Hobel
pos1t1ve cells were utilized. None of the injected compounds appeared to decrease the induced antibody titer in the dosages given (Tables V and VI). Preliminary human studies. Since glucosamine is a naturally occurring substance in the human, 10 per cent glucosamine in distilled water was prepared under sterile conditions in an infusion bottle. At this writing, only one patient has received the compound.
Case report She is a 24-year-old para 2-0-0-2, who is type 0 Rh negative. Her husband is type A Rh positive. A first child, born elsewhere in 1958 was normal. A second child, born elsewhere, in 1960 after an induced labor at 36 weeks' gestation required three exchange transfusions but survived. Both children are 0, Rh positive. When seen for the first time, at 23 weeks' gestation, with her third pregnancy, the Rh antibody titer was 1: 1 in saline, 1: 16 in albumin, and 1 :32 Coombs. The pregnancy was otherwise uncomplicated. At 27 weeks' gestation, 900 ml. of 10 per <.:ent D-glucosamine was infused intravenously over a 3 hour period. There were no side effects apart from the anticipated diuresis. Antibody titers before and immediately after the infusion as well as the effects of the infused solution added to the pretreatment serum in vitro are recorded in Table VII. The 10 per cent glucosamine in vitro was effective as anticipated. In vivo, there appeared to be no effect on saline antibodies. While the effect on the Race score measured in albumin and with Coombs serum appears minimal it was reproducible, and the last agglutination was consistently visible one tube earlier in the treated serum (decreased titer). The possibility that this was purely an effect of hemodilution must be considered. However, urinary output over the 3 hours approximately equalled the volume of solution infused. One week later a minimal Race score reduction persisted. The patient was delivered of a stillborn hydropic infant at 31 weeks' gestation. Comment
The effects of the test compounds are reproducible and appear specific for the Rh as opposed to the i\BO or j\fl'! blood group systems.
\rn.
J.
:\1ay l, J%{ Ohst. & Gmc'-
The original postulation for these observations was that the ribonucleic acid derivatives were similar to the terminal chemical groups of the erythrocyte antigen structures. Subsequent observations by Hackel and Smolker8 on the effects of ribonuclease in decreasing agglutinability of RBC would tend to confirm this impression. It is known that glucosamine and galactosamine are components of polysaccharide blood group antigens. The studies of Ricci and associates 17 in which the test compounds were added to various erythrocyte suspensions, and antisera were added subsequently, generally demonstrated no evidence of inhibition. This would appear to indicate that the reaction was between compound and antisera not compound and red cell. It should also be observed that the compounds found effective are "electron rich," with free phosphate and amine radicals. Thus, the antibody inhibition may be on a stero chemical basis, on the basis of electron transfer, or both. Further studies on the nature of the reaction, utilizing sugar phosphate derivatives and other compounds at varying temperatures, are in progress. \Ve believe the amine radical to be the key factor. The infusion of 10 per cent glucosaminc should have achieved approximately a 2 per cent concentration in the subject, a level at which only minimal inhibition has been observed in vitro. The apparent immediate results on treated serum were of about the anticipated magnitude. Although minimal, they were reproducible on repeated determinations with the microscopic readings made by separate observers given the slides randomly. Summary
o!
. 1. Of fou~ ~e:hods preventing or treatIng Rh sens1t1zat1on, thlS study concerns the area of inhibition of maternal antibodies. 2. Four ribonucleic acid derivatives inhibit the reaction of test and maternal sera containing complete and incomplete Rh antibodies. 3. Two sugar-amines, glucosamine, and
Volume 89 Number 1
galactosamine have also been demonstrated to inhibit complete and incomplete Rh antibodies in test and maternal sera. 4. Specific Rh antibodies could not be induced in rabbits by the techniques em-
Inhibition of Rh antibodies 51
ployed, nor were the antibody levels produced affected by multiple small injections of the experimental compounds. 5. Preliminary use of D-glucosamine m an Rh-sensitized patient is reported.
REFERENCES
1. Belkin, R. B., and Weiner, A. S.: Proc. Soc. Exper. Bioi. & Med. 56: 214, 1944. 2. Boyd, W. C., McMaster, M. H., and Wasczenko-Zacharczenko, E.: Nature 184: 989, 1959. 3. Carter, B. B., Williamson, A. C., Loughrey, ]., and Ingram, C. H., Jr.: AM. J OBsT. & GYNEC. 72: 655, 1956. 4. Evans, T. N.: AM. J. 0BsT. & GYNEC. 72: 312, 1956. 5. Finn, R., Clarke, C. A., Donohoe, W. T. A., McConnell, R. B., Sheppard, P. M., Lehane, D., and Kulke, W.: Brit. M. J. 1: 1486, 1961. 6. Gainey, H. L., Nicolay, K. S., Keeler, J. E., and Doyle, M. E.: Obst. & Gynec. 3: 141, 1954. 7. Hackel, E., Smolker, R. E., and Fenske, S. A.: Vox Sanguinus 3: 402, 1958. 8. Hackel, E., and Smolker, R. E.: Nature 187: 1037, 1960.
9. Hackel, E., and Spoylar, K. S.: Vox Sanguinus 5: 517, 1960. 10. Hubinot, P.: Nature 161: 642, 1948. 11. Hunter, 0. B.: J. A. M. A. 154: 905, 1954. 12. Jacobs, W. M.: Surg. Gynec. & Obst. 110: 33, 1960. 13. Keenan, H. W., and Pearse, W. H.: AM. J. 0BST. & GYNEC. 86: 1096, 1963. 14. Lubinski, H. H., and Portnuff, J. C.: J. Lab. & Clin. Med. 32: 178, 1947. 15. Maloney, W. C.: AM. J. 0BST. & GYNEC. 60: 616, 1950. 16. Race, R. R., and Sanger, R.: Blood Groups in Man, ed. 2, Oxford, 1954, Blackwell Scientific Publications, p. 275. 17. Ricci, C., Coscia, C., Pileri, A., and Gavosto, F.: Bull. Soc. Ita!. Bioi. Sper. 36: 233, 1960. 42nd and Dewey Ave. Omaha, Nebraska
Discussion Galveston, Texas. The majority of cases of Pill sensitization (isoimmunization) can be recognized by their antibody formation in the mother prenatally, but we see exceptions frequently enough to lead us to doubt the value of antibody determinations in either a positive or negative way. For example, the highest titer we have ever recorded in our hospital's blood bank laboratory has been in a case of anamnestic reaction. One must be cautious in interpreting minor changes in titer in vivo during pregnancy. Our work in this field for the last several years showed that there was a paucity of data available in the literature as to what happens in the untreated patient on a day-to-day or week-to-week basis. We have tried to accumulate some of this information and have been able to show steady drops in titer on a day-to-day basis in untreated patients as well as no change and rises. Based initially on some observations of Ritzman and Levin from our university of the effects DR. CHARLES PowELL, }R.,
of certain mercaptans on macroglobulins in cold agglutinin disease \ve noted the similarity in the antibody structures to the Rh antibody complex. The complex Rh (saline) antibodies are found in the macroglobulin fraction (the B2 M or 19S gamma-1) and the incomplete (albumin and Coombs) are found in the B2 -A and gamma globulin fraction ( 7S). These are the so-called immunoglobulins. In addition to immunological differences, there are physicochemical differences between these immunoglobulins. The molecular weight of a macroglobulin is around 1 million, and this is thought to be the reason that the saline antibody is not found in the baby, as these large particles will not cross the placenta. We have found them in amniotic fluid, however. In contrast, the B2 A gamma globulin molecular weights are around 160,000, and these readily traverse the "placental barrier" and thus are thought to be the actual cause of the isoimmunization. The active sulfhydryl (SH) groups of mer-
52
Pearse and Hobel
captans can split these large-sized macroglobulins by their active hydrogen into smaller globulins that are possibly inert. Work by Diamond and co-workers has also shown suggestive evidence that the incomplete and complete Rh antibodies may mask the presence of one or another. As the compounds used by the authors also have available "H" ions, they may be exerting a similar type of physicochemical reaction. We have been able to show in vitro and in vivo reduction of both complete and incomplete Rh antibodies with the use of several mercaptans-penicillin (part of which undergoes transformation in fresh serum to penicillamine), d, 1 penicillamine (d, I dimethylcysteine) and vitamin B6 SH (Me rca ptopyridoxine) . All titer results discussed are done in our research lab, using the same technician, clean pipettes for each dilution, with the sante fresh Rh-positive test cells and same batch number of test serum being used each determination, run in duplicate, with positive and negative controls done on each run. If these conditions are not followed, gross misinterpretations can be made of results, especially if one is trying to detect qualitative changes on each test dilution ( 1 to 4+ or Race scale values) and the value of changes of less than a 2 tube dilution difference of actual titer and points are probably not significant. Few people continue to work with albumin antibodies because of appearance of prozones, and this is one minor criticism I have of Dr. Pearse's work. Even in small concentrations B6 SH has been effective in vitro in reducing Rh titers. In concentrations of as little as 1 mg. per cubic centimeter we have gotten a 4 tube dilution change in the saline titer and a 1 tube reduction in Coombs. Five milligrams per cubic centimeter produced as much as a 5 tube change in saline titer and a 2 tube change in Coombs titer. We recently had one patient upon whom it appeared warranted to attempt antenatal reduction of antibodies in an effort to perhaps circumvent a seriously affected infant, and to gain information as to the effectiveness of this approach. When first seen, the patient had had several affected infants, was late in pregnancy, with a high saline and Coombs titer which had been rising, according to her referring physician. She was found to be sensitized to both C and D. Her husband was heterozygous, howevl'r. The results of our efforts of in vivo daily intravenous administration of 1,500 mg. B0 SH produced a 2 tube drop in titer of both saline and Coombs antibodv <'nd points and when Race score values are used,
Am.
J.
May I, 1964 Obst. & Gynec.
the change was even more impressive as the Race scale value in Coombs titer dropped from 70 to 32 and saline values from 28 to 7. This course of treatment was carried out over 10 days. This case was helpful in showing that we could reduce the titer, but it also points out the fallacy of titer determinations, i.e., this woman had an Rhnl'gative unaffected infant. Penicillin and penicillamine must be 111 stronger concentrations to exhibit an effect m vitro, but we have gotten a good reduction in antibody titers in one case in vivo. This patient was treated for 50 days with 1,000 mg. of d, l penicillamine, starting around her twenty-sixth week, had a homozygous husband, with a very bad Rh obstetrical history of affected infants. Again, a very good response was obtained with regards to reduction in titer values, but a hydropic stillbirth infant was the rl'suit. It should be stated that this baby was immunized against both C and D. In only 1 week the titer was progressively dropping with an over-all change in Race scores, in Coombs titer from S:i to a low of 8, and saline changed from 14 to 3. Until we have more control data, minor changes of titer values will have to be int<~r preted cautiously. Drs. Pearse and Robel state in their paper that sera with saline antibodies from sensitized patients were of generally low magnitude and usually disappeared even though the sera were frozen. This has not been our experience. vVe have been able to keep sera for several years in a standard deep freeze with several thawings without appreciable change in the titer values. Saline antibodies are heat labile and disappear in around 14 to 16 days at 3 7° C., and generally disappear after delivery, but in the case we discussed with penicillamine treatmrnt the salirw titer was positive again after two years and no intervening pregnancy or transfusions. We cannot explain this finding as it should not occur, going by previous experience. We also feel that it is worth while to contimw efforts at desensitizing these patients by such an approach, but to find safl' compounds to use and, even more important, the proper patients that -" __ L __ >UlH cumpounas may oe grven to with hopes of getting a clear-cut answer to the clinical effectiveness. Maloney has aptly stated these criteria which are summarized here. 1. Homozygous husbands ( '
1
1
•
•
~
•
Volume 89 Number 1
3. Hx of loss of at least the prior pregnancy (stillborn or hydrops) due to hemolytic disease. 4. Therapy started early, preferably in the first trimester. Unless we choose patients in this category to study and treat prospectively, doubt will remain in everyone's eyes as to the results obtained. We have found only one case every year and a half that meets these criteria and this makes it a slow process to accumulate a large series. Once we are able to learn more about the basic nature of antibody-antigen reactions, their structures, etc., we should be able to more effectively control the ravages of erythroblastosis. I feel the ultimate answer is going to be found in work like that which has been presented here today, and I urge the authors to continue their endeavors. DR. PEARSE (Closing). I would like to make four brief comments in general concurrence with what Dr. Powell has said. First of all, I could not agree more with his opinion about the value of antibody titers. I believe that if they are posi-
Inhibition of Rh antibodies
53
tive the patient is sensitized to the Rh factor and that value changes up and down are of little significance, and these certainly occur commonly as an anamnestic reaction. I agree also with the comments about albumin antibodies and problems that exist with prozone formation. A table which appears in the formal paper illustrates a very definite prozone effect in albumin, and we continue to determine the albumin antibodies simply for completeness sake. It has been our experience that the saline antibodies have disappeared on freezing, but this has been only the 19S antibodies. We have found the titer of other antibodies not at all affected by prolonged freezing, and some of our sera have been kept for 1;/2 to 2 years. Finally, I agree that both these studies are preliminary but interesting. The final answer to this problem is going to be successful treatment of patients who have erythroblastotic infants and, in the meantime, our studies and Dr. Powell's studies will help to teach us more about the antigen-antibody reaction.
S I N c E the etiology of the long puzzling condition, erythroblastosis fetalis, was discovered in 1940, many attempts at modification or elimination of the disease have been made. These treatment attempts basically fall into four categories: ( 1) Initial sensitization of the mother may be prevented by protecting placental integrityl 2· 13 or avoiding intrauterine trauma at delivery 6 ; (2) presumed or demonstrated sensitization may be blocked by injection of antibody to neutralize fetal cells which have gained entrance to the maternal circulation 5 ; ( 3) attempts may be made to neutralize or inactivate demonstrated circulating antibody in the maternal circulation before significant amounts cross the placenta (previous efforts in this area include the use of Rh hapten, 3 cortisoP 1 and ethylene disulfonate) 15 and (4) the affected infant is accepted, perhaps delivered preterm, and, if necessary, exchange transfusion is carried
fate, adenylic acid, and uridylic acid in 2 per cent concentration at pH 6.8 to be inhibitors of an ti-D (Rho) and this was later extended to anti-c (hr') and anti-e (hr") .9 A large series of other compounds tested, including sugars, amino acids, desoxyribonucleic acid (DNA) derivatives, purines, and pyrimidines showed no similar effect. The report of Boyd, McMaster, and Wasczenko-Zacharczenko2 demonstrated inhibition of anti-D by high concentrations (0.1 to 0.2 molar) of "unnatural sugars," especially !-glucose, 1mannose, and d-gulose. Since the saline antibodies, whose inhibition was demonstrated by these authors, are not transplacentally transmitted, it seemed desirable to extend these studies to the albumin reacting or "incomplete" antibodies ( 7S) which are the pathogenic agents in erythroblastosis fetalis. Further, while such in vitro studies give valuable evidence concerning the chemical structure of red cell surface antigens, and antigen-antibody reactions, it also seemed desirable to extend the investigation to in vivo studies, with the ultimate aim of aiding the problem of Rh isoimmunization in the human.
---.&- 4 UUL.
The present study considers a new application of the third approach. Inhibition of antiRh antibodies (as well as anti-Lutheran) by ribonucleic acid derivatives was first reported by Hackel, Smolker, and Fenske. 7 These studies, carried out entirely in saline media, demonstrated cytidylic acid, cytidine hemisul-
Materials and methods
Inhibition titers. Compounds to be tested for antibody inhibiting effect were prepared in varying concentrations from 1 to 20 per cent in 0.2M phosphate buffer. Tenth normal potassium hydroxide was used to adjust pH to either 6.8 or 7.4. Standard anti-D sera of the saline reacting and albumin reacting types as well as patients' sera were serially diluted with normal
From the Department of Obstetrics and Gynecology, University of Nebraska College of Medicine. Supported in part by Grant No. 2G-631 from the United States Public Health Service. Presented at the Thirty-first Annual · Meeting of the Central Association of Obstetricians and Gynecologists, Denver, Colorado, Sept. 12-14, 1963.
46
Volume 39
Inhibition of Rh antibodies
.!\umber 1
Table I. Example of the Race agglutination scale
Table II. Sera inhibited with ribonucleic acid derivatives ( 8 to 20 per cent) in 11 patients
Agglu- 1: 1 1:2 1:4 1:8 1: 16 1:32 1:64 1: 128 tination 10 10 10 8 5 2 0 0 Saline anti-D serum + 2 per cent suspension 0+ red blood cells in saline Total score
= 45
Average
Saline Albumin Coombs
Control Race score
Treated Race score
0 24
0
44
22
Control A and B serum
saline or type AB serum through ten tubes, the volume in each tube being 0.1 ml. Onetenth milliliter of the test solution was added to each tube and the mixture incubated at 37° C. for 30 minutes. One-tenth milliliter of type 0 Rh-positive red blood cells in 2 per cent suspension in either saline or protein media as appropriate, was added to each tube and incubated at 37° C. for 60 minutes. Positive controls utilizing saline instead of the experimental compound and negative controls utilizing type 0 Rhnegative red blood cell suspensions were routinely employed. After 60 minutes, the tubes were centrifuged at 1,000 r.p.m. for 1 minute, the cells resuspended by gentle tapping and the tube contents examined microscopically. Agglutination was graded by the Race scale/ 6 ranging from a score of 10 for agglutination clearly visible to the naked eye, to 8 for large microscopic agglutinates, 5 for average size agglutinates, 3 for small agglutinates, 2 for the smallest definite agglutinates seen microscopically, and 0 for no evidence of agglutination (Table I). In most instances, two examiners graded the slides independently. When no agglutination was demonstrated in the saline system, the cells were washed three times in normal saline with low speed centrifugation, 0.1 ml. of Coombs antihuman globulin was added, the tube was centrifuged at 500 r.p.m. for 1 minute and re-examined microscopically as noted above. The same donor was used as a source of type 0 Rh-positive cells throughout the nucleic acid derivative tube and animal studies. Varying donors were employed for the remainder of the study.
47
2
= no inhibition
Table III. Compounds* inhibiting albumin antibodies in the serum of Patient M. S. Race score Saline
Albu-1 min Coombs
Control 0 38 62 Glucose 10 per cent 20 60 Citric acid 10 per cent 10 66 Succinic acid 10 per cent 0 58 50 Reticulose 8 60 0 Glycine 10 per cent Ethylene diacetate 10 -hemolyzed cellsper cent *Tests of these compounds against saline Anti-D test sera demonstrated no inhibitory effect.
Erythrocyte antigen for rabbit injection. Rh-positive red cell antigens for rabbit injection were prepared in two fashions: 1. Belkin and Weiner1 demonstrated that suspensions of cell stromata retained their antigenic properties. Type 0, Rh genotype Dee/deE red blood cells were washed in isotonic saline until a clear supernatant was obtained. The cells were then lysed in distilled water with three washings, the ghost cells centrifuged and mixed with normal saline to provide a 1/10 dilution. 2. Lubinski and Portnuff14 pointed out that heating Rh-positive cells suspended in saline to 56° C. rendered the cells Rh negative. Hubinot10 then demonstrated that an Rh substance was present in the supernatant saline by inhibition studies. The same type 0, Rh genotype Dee/deE red blood cells were washed in isotonic saline until clear, and with an equal amount of saline added the mixture was adjusted to pH 7.2 by adding
48
Pearse and Hobel
..\.111.
INHIBITION OF SALINE ANTI-D TEST SERA
§ ... g
15
v
;::
~
Results
25 . , ,
20
\
' • \\
10
·'\.-..
~CONTROL
May 1, IY64 Obst. & GYnec.
Control rabbits received the buffer solution alone as a subcutaneous injection over a '1 ""'! , • .l snnuar 1 uay penuu.
~=l ~
J.
10
2
DRUG CONCENTRATION
0
12
IPIR
14
16
18
20
CENT)
= ADENYLIC ACID
· = CYTIDINE HEMISULFATE Fig. 1.
O.lN KOH. The cells were then heated for 17 minutes at 56° C., centrifuged, and the supernatant fluid recovered. This was mixed with normal saline to provide a 1/10 dilution. Animal studies. Control studies of naturally occurring agglutinins to 0 Rh-positive human red cells in rabbit serum were carried out. Six rabbits were then injected subcutaneously with i mi. of cell stromata preparation on days 1, 2, 3, and 8, 9 and 10. Six rabbits were sensitized in the same fashion with the heated supernatant preparation. Three control rabbits were injected in the same fashion with a cell stromata preparation of type 0, Rh-negative cells. A second titer level was obtained on day 7, a third on day 17, and a fourth on day 29. Specificity of the induced antibodies in the rabbit's sera was evaluated by selective absorption and tests against a panel of known human cells. As noted below, specificity could not be demonstrated. Antibody inhibition or inactivation was attempted by injecting subcutaneously 1.0 ml. of a test compound daily on days 32 to 39 and again determining antibody levels on day 40.
1. The four RNA derivatives originally proposed (adenylic acid, uridylic acid, cytidylic acid and cytidine hemisulfate) were tested against commercial an ti-D (Rho) serum and against individual patient sera. Extensive initial testing in saline revealed minimal effect in 2 per cent concentration and this was evident only in serum dilutions. Fig. 1 illustrates results characteristic of the four compounds. The adenylic acid 5' did not maintain effectiveness more than 24 hours unless frozen, whereas the other compounds retained activity as long as 6 months at 4° C. In protein media, and utilizing appropriate cell suspensions in AB serum all four cornpounds were equally effective in inhibiting "albumin" antibodies. However, the "incomplete" antibodies demonstrated by Coombs antihuman globulin persisted, although reduced, when saline ("complete") antibodies were absent. Results on 11 serum specimens from patients vvith kno\vn erythroblastotic infants are outlined in Table II. Saline titers in sera obtained from sensitized patients wPn' generally of low magnitude ( 1 : 1 or 1 : 2) and usually disappeared even though the sera were frozen. Compounds were also tested against anti-A, anti-B, anti-M, and anti-?\ test sera. There was no inhibiting effect on agglutination in any instance. 2. A search was begun for more effective inhibiting compounds. Among those tested were glucuronic acid ( 10 per cent), glucose ( 10 and 20 per cent) , sucrose ( 10 and 20 per cent), citric acid ( 10 per cent) , succinic acid ( 10 per cent), reticulose (a "nucleic acid complex"), and glycine ( 10 per cent i. Surprisingly, while none of these compounds were effective against saline anti-D test sera, nor in inhibiting Coombs antibodies in patient serum, they all seemed to have an inhibitory effect in albumin media (Table III!. Ethylene diacetate was tested since a similar compound, ethylene disulfonate, had been suggested as an antibody inhibitor in minute
Volume 89 Number 1
Inhibition of Rh antibodies
49
Table IV. Test sera inhibited with sugar amine derivatives* 10 per cent
I
Race score control
10 per cent glucosamine
n-acetyl glucosamine
galactosamine
10 per cent n-acetyl galactosamine
33 17
0 0 0
17 2
0
20
0
2
Saline Albumin Coombs
10 per cent
0
*Control A and B serum = no inhibition. Equivalent results at pH 6.8 and 7.4.
Table V. Rabbit saline titers-Race score
Titer
Rabbits 1 through 6
I (control)
10 25 20
6 16 36 28
4 8 21 19
19 25 10
26 42 13
12
1-
II (day 7) III (day 17)
IV (day 29) v (day 40) DG* CH* CON*
Rabbits Rabbits 7 through 13 through 12 15
=
4
13
D-glucosamine, 10 per cent; CH == cytidine *DG buffer control. hemisulfate, 15 per cent; and CON
=
Table VI. Rabbit albumin titers-Race score
Titer
Rabbits 1 through 6
I (control)
II (day 7) III (day 17
IV (day 29) V (day 40) DG* CH* CON*
I7Rabbits through 12
Rabbits 13 through 15
6
9
4
11 23 20
19 36 27
14 21 18
15 18 18
11 18
6 4
3
14
*DG = D-glucosamine, 10 per cent; CH = cytidine buffer control. hemisulfate, 15 per cent; and CON
=
Table VII. Ten per cent glucosamine in patient
quantitiesY In a 10 per cent solution, all red cells were hemolyzed. A series of compounds was discovered, however, which appeared to inhibit saline, albumin, and Coombs antibodies more effectively than the ribonucleic acid derivatives. These were glucosamine, N-acetyl glucosamine, galactosamine, n-acetyl galactosamine, and pyridoxal phosphate. The acetyl derivatives were found less effective and not extensively tested (Table IV). Further multiple studies with anti-D and anti-C commercial sera and patient sera confirmed the effectiveness of glucosamine and galactosamine in all three antibody systems. Animal studies. Animal results were disappointing. Naturally occurring antibodies to human 0 Rh positive cells were present in an average Race score of 5 in saline and 7 in protein media. Rabbits were readily immunized to the cell stromata preparation and supernatant fluid preparation, but specificity of the antibodies could not be demonstrated with the use of panel cells or by selective absorption. Immunization with the cell stromata preparation produced equivalent sensitization utilizing both Rh-positive and Rh-negative cells. The heated supernatant preparation produced higher antibody titers, but only Rh-
J- M.
Control serum Rh - cells
Rh + cells (titer)
Glucosamine tube Rh + cells (titer)
Posttreatment sera Rh + cells (titer)
0
33 ( 1 :4) 75 (1:64) 92 (1:256)
10 (0) 18 ( 1: 1) 32 (1:2)
32 ( 1: 4) 68 (1:32) 85 (1:128)
Saline Race score Albumin
0
Coo1nbs Race score
0
50 Pearse and Hobel
pos1t1ve cells were utilized. None of the injected compounds appeared to decrease the induced antibody titer in the dosages given (Tables V and VI). Preliminary human studies. Since glucosamine is a naturally occurring substance in the human, 10 per cent glucosamine in distilled water was prepared under sterile conditions in an infusion bottle. At this writing, only one patient has received the compound.
Case report She is a 24-year-old para 2-0-0-2, who is type 0 Rh negative. Her husband is type A Rh positive. A first child, born elsewhere in 1958 was normal. A second child, born elsewhere, in 1960 after an induced labor at 36 weeks' gestation required three exchange transfusions but survived. Both children are 0, Rh positive. When seen for the first time, at 23 weeks' gestation, with her third pregnancy, the Rh antibody titer was 1: 1 in saline, 1: 16 in albumin, and 1 :32 Coombs. The pregnancy was otherwise uncomplicated. At 27 weeks' gestation, 900 ml. of 10 per <.:ent D-glucosamine was infused intravenously over a 3 hour period. There were no side effects apart from the anticipated diuresis. Antibody titers before and immediately after the infusion as well as the effects of the infused solution added to the pretreatment serum in vitro are recorded in Table VII. The 10 per cent glucosamine in vitro was effective as anticipated. In vivo, there appeared to be no effect on saline antibodies. While the effect on the Race score measured in albumin and with Coombs serum appears minimal it was reproducible, and the last agglutination was consistently visible one tube earlier in the treated serum (decreased titer). The possibility that this was purely an effect of hemodilution must be considered. However, urinary output over the 3 hours approximately equalled the volume of solution infused. One week later a minimal Race score reduction persisted. The patient was delivered of a stillborn hydropic infant at 31 weeks' gestation. Comment
The effects of the test compounds are reproducible and appear specific for the Rh as opposed to the i\BO or j\fl'! blood group systems.
\rn.
J.
:\1ay l, J%{ Ohst. & Gmc'-
The original postulation for these observations was that the ribonucleic acid derivatives were similar to the terminal chemical groups of the erythrocyte antigen structures. Subsequent observations by Hackel and Smolker8 on the effects of ribonuclease in decreasing agglutinability of RBC would tend to confirm this impression. It is known that glucosamine and galactosamine are components of polysaccharide blood group antigens. The studies of Ricci and associates 17 in which the test compounds were added to various erythrocyte suspensions, and antisera were added subsequently, generally demonstrated no evidence of inhibition. This would appear to indicate that the reaction was between compound and antisera not compound and red cell. It should also be observed that the compounds found effective are "electron rich," with free phosphate and amine radicals. Thus, the antibody inhibition may be on a stero chemical basis, on the basis of electron transfer, or both. Further studies on the nature of the reaction, utilizing sugar phosphate derivatives and other compounds at varying temperatures, are in progress. \Ve believe the amine radical to be the key factor. The infusion of 10 per cent glucosaminc should have achieved approximately a 2 per cent concentration in the subject, a level at which only minimal inhibition has been observed in vitro. The apparent immediate results on treated serum were of about the anticipated magnitude. Although minimal, they were reproducible on repeated determinations with the microscopic readings made by separate observers given the slides randomly. Summary
o!
. 1. Of fou~ ~e:hods preventing or treatIng Rh sens1t1zat1on, thlS study concerns the area of inhibition of maternal antibodies. 2. Four ribonucleic acid derivatives inhibit the reaction of test and maternal sera containing complete and incomplete Rh antibodies. 3. Two sugar-amines, glucosamine, and
Volume 89 Number 1
galactosamine have also been demonstrated to inhibit complete and incomplete Rh antibodies in test and maternal sera. 4. Specific Rh antibodies could not be induced in rabbits by the techniques em-
Inhibition of Rh antibodies 51
ployed, nor were the antibody levels produced affected by multiple small injections of the experimental compounds. 5. Preliminary use of D-glucosamine m an Rh-sensitized patient is reported.
REFERENCES
1. Belkin, R. B., and Weiner, A. S.: Proc. Soc. Exper. Bioi. & Med. 56: 214, 1944. 2. Boyd, W. C., McMaster, M. H., and Wasczenko-Zacharczenko, E.: Nature 184: 989, 1959. 3. Carter, B. B., Williamson, A. C., Loughrey, ]., and Ingram, C. H., Jr.: AM. J OBsT. & GYNEC. 72: 655, 1956. 4. Evans, T. N.: AM. J. 0BsT. & GYNEC. 72: 312, 1956. 5. Finn, R., Clarke, C. A., Donohoe, W. T. A., McConnell, R. B., Sheppard, P. M., Lehane, D., and Kulke, W.: Brit. M. J. 1: 1486, 1961. 6. Gainey, H. L., Nicolay, K. S., Keeler, J. E., and Doyle, M. E.: Obst. & Gynec. 3: 141, 1954. 7. Hackel, E., Smolker, R. E., and Fenske, S. A.: Vox Sanguinus 3: 402, 1958. 8. Hackel, E., and Smolker, R. E.: Nature 187: 1037, 1960.
9. Hackel, E., and Spoylar, K. S.: Vox Sanguinus 5: 517, 1960. 10. Hubinot, P.: Nature 161: 642, 1948. 11. Hunter, 0. B.: J. A. M. A. 154: 905, 1954. 12. Jacobs, W. M.: Surg. Gynec. & Obst. 110: 33, 1960. 13. Keenan, H. W., and Pearse, W. H.: AM. J. 0BST. & GYNEC. 86: 1096, 1963. 14. Lubinski, H. H., and Portnuff, J. C.: J. Lab. & Clin. Med. 32: 178, 1947. 15. Maloney, W. C.: AM. J. 0BST. & GYNEC. 60: 616, 1950. 16. Race, R. R., and Sanger, R.: Blood Groups in Man, ed. 2, Oxford, 1954, Blackwell Scientific Publications, p. 275. 17. Ricci, C., Coscia, C., Pileri, A., and Gavosto, F.: Bull. Soc. Ita!. Bioi. Sper. 36: 233, 1960. 42nd and Dewey Ave. Omaha, Nebraska
Discussion Galveston, Texas. The majority of cases of Pill sensitization (isoimmunization) can be recognized by their antibody formation in the mother prenatally, but we see exceptions frequently enough to lead us to doubt the value of antibody determinations in either a positive or negative way. For example, the highest titer we have ever recorded in our hospital's blood bank laboratory has been in a case of anamnestic reaction. One must be cautious in interpreting minor changes in titer in vivo during pregnancy. Our work in this field for the last several years showed that there was a paucity of data available in the literature as to what happens in the untreated patient on a day-to-day or week-to-week basis. We have tried to accumulate some of this information and have been able to show steady drops in titer on a day-to-day basis in untreated patients as well as no change and rises. Based initially on some observations of Ritzman and Levin from our university of the effects DR. CHARLES PowELL, }R.,
of certain mercaptans on macroglobulins in cold agglutinin disease \ve noted the similarity in the antibody structures to the Rh antibody complex. The complex Rh (saline) antibodies are found in the macroglobulin fraction (the B2 M or 19S gamma-1) and the incomplete (albumin and Coombs) are found in the B2 -A and gamma globulin fraction ( 7S). These are the so-called immunoglobulins. In addition to immunological differences, there are physicochemical differences between these immunoglobulins. The molecular weight of a macroglobulin is around 1 million, and this is thought to be the reason that the saline antibody is not found in the baby, as these large particles will not cross the placenta. We have found them in amniotic fluid, however. In contrast, the B2 A gamma globulin molecular weights are around 160,000, and these readily traverse the "placental barrier" and thus are thought to be the actual cause of the isoimmunization. The active sulfhydryl (SH) groups of mer-
52
Pearse and Hobel
captans can split these large-sized macroglobulins by their active hydrogen into smaller globulins that are possibly inert. Work by Diamond and co-workers has also shown suggestive evidence that the incomplete and complete Rh antibodies may mask the presence of one or another. As the compounds used by the authors also have available "H" ions, they may be exerting a similar type of physicochemical reaction. We have been able to show in vitro and in vivo reduction of both complete and incomplete Rh antibodies with the use of several mercaptans-penicillin (part of which undergoes transformation in fresh serum to penicillamine), d, 1 penicillamine (d, I dimethylcysteine) and vitamin B6 SH (Me rca ptopyridoxine) . All titer results discussed are done in our research lab, using the same technician, clean pipettes for each dilution, with the sante fresh Rh-positive test cells and same batch number of test serum being used each determination, run in duplicate, with positive and negative controls done on each run. If these conditions are not followed, gross misinterpretations can be made of results, especially if one is trying to detect qualitative changes on each test dilution ( 1 to 4+ or Race scale values) and the value of changes of less than a 2 tube dilution difference of actual titer and points are probably not significant. Few people continue to work with albumin antibodies because of appearance of prozones, and this is one minor criticism I have of Dr. Pearse's work. Even in small concentrations B6 SH has been effective in vitro in reducing Rh titers. In concentrations of as little as 1 mg. per cubic centimeter we have gotten a 4 tube dilution change in the saline titer and a 1 tube reduction in Coombs. Five milligrams per cubic centimeter produced as much as a 5 tube change in saline titer and a 2 tube change in Coombs titer. We recently had one patient upon whom it appeared warranted to attempt antenatal reduction of antibodies in an effort to perhaps circumvent a seriously affected infant, and to gain information as to the effectiveness of this approach. When first seen, the patient had had several affected infants, was late in pregnancy, with a high saline and Coombs titer which had been rising, according to her referring physician. She was found to be sensitized to both C and D. Her husband was heterozygous, howevl'r. The results of our efforts of in vivo daily intravenous administration of 1,500 mg. B0 SH produced a 2 tube drop in titer of both saline and Coombs antibodv <'nd points and when Race score values are used,
Am.
J.
May I, 1964 Obst. & Gynec.
the change was even more impressive as the Race scale value in Coombs titer dropped from 70 to 32 and saline values from 28 to 7. This course of treatment was carried out over 10 days. This case was helpful in showing that we could reduce the titer, but it also points out the fallacy of titer determinations, i.e., this woman had an Rhnl'gative unaffected infant. Penicillin and penicillamine must be 111 stronger concentrations to exhibit an effect m vitro, but we have gotten a good reduction in antibody titers in one case in vivo. This patient was treated for 50 days with 1,000 mg. of d, l penicillamine, starting around her twenty-sixth week, had a homozygous husband, with a very bad Rh obstetrical history of affected infants. Again, a very good response was obtained with regards to reduction in titer values, but a hydropic stillbirth infant was the rl'suit. It should be stated that this baby was immunized against both C and D. In only 1 week the titer was progressively dropping with an over-all change in Race scores, in Coombs titer from S:i to a low of 8, and saline changed from 14 to 3. Until we have more control data, minor changes of titer values will have to be int<~r preted cautiously. Drs. Pearse and Robel state in their paper that sera with saline antibodies from sensitized patients were of generally low magnitude and usually disappeared even though the sera were frozen. This has not been our experience. vVe have been able to keep sera for several years in a standard deep freeze with several thawings without appreciable change in the titer values. Saline antibodies are heat labile and disappear in around 14 to 16 days at 3 7° C., and generally disappear after delivery, but in the case we discussed with penicillamine treatmrnt the salirw titer was positive again after two years and no intervening pregnancy or transfusions. We cannot explain this finding as it should not occur, going by previous experience. We also feel that it is worth while to contimw efforts at desensitizing these patients by such an approach, but to find safl' compounds to use and, even more important, the proper patients that -" __ L __ >UlH cumpounas may oe grven to with hopes of getting a clear-cut answer to the clinical effectiveness. Maloney has aptly stated these criteria which are summarized here. 1. Homozygous husbands ( '
1
1
•
•
~
•
Volume 89 Number 1
3. Hx of loss of at least the prior pregnancy (stillborn or hydrops) due to hemolytic disease. 4. Therapy started early, preferably in the first trimester. Unless we choose patients in this category to study and treat prospectively, doubt will remain in everyone's eyes as to the results obtained. We have found only one case every year and a half that meets these criteria and this makes it a slow process to accumulate a large series. Once we are able to learn more about the basic nature of antibody-antigen reactions, their structures, etc., we should be able to more effectively control the ravages of erythroblastosis. I feel the ultimate answer is going to be found in work like that which has been presented here today, and I urge the authors to continue their endeavors. DR. PEARSE (Closing). I would like to make four brief comments in general concurrence with what Dr. Powell has said. First of all, I could not agree more with his opinion about the value of antibody titers. I believe that if they are posi-
Inhibition of Rh antibodies
53
tive the patient is sensitized to the Rh factor and that value changes up and down are of little significance, and these certainly occur commonly as an anamnestic reaction. I agree also with the comments about albumin antibodies and problems that exist with prozone formation. A table which appears in the formal paper illustrates a very definite prozone effect in albumin, and we continue to determine the albumin antibodies simply for completeness sake. It has been our experience that the saline antibodies have disappeared on freezing, but this has been only the 19S antibodies. We have found the titer of other antibodies not at all affected by prolonged freezing, and some of our sera have been kept for 1;/2 to 2 years. Finally, I agree that both these studies are preliminary but interesting. The final answer to this problem is going to be successful treatment of patients who have erythroblastotic infants and, in the meantime, our studies and Dr. Powell's studies will help to teach us more about the antigen-antibody reaction.