CURRENT COMMENT The Problem of Standardization in Rheumatoid Arthritis Serology By S. BOZS~KY T HE CLINICAL SIGNIFICANCE of the so-called rheumatoid factors is not completely understood. Some of the contradictory data may be due partly to the wide variety of serologic methods adapted at different laboratories. In order to elucidate this problem, at the Xth International Congress of Rheumatology in Rome a world-wide comparative study was initiated. Freezedried samples of a “positive” and a “negative” serum pool, collected from 50 rheumatoid arthritis patients and 50 normal subjects, respectively, were distributed among the participants.* About 30 serologists from 14 countries were kind enough to take over the task of performing the comparative investigations. To date, 26 protocols have been returned, and have afforded the following information. The sera were investigated by the sensitized sheep cell reaction (SSC test) at 19 laboratories, and by the latex fixation reaction (LF-I1 test) in 14 cases. Both reactions were simultaneously performed, however, in only nine instances. In addition to, or instead of these two methods, in some cases the sensitized human cell reaction ( SHC test), tanned sheep cell reaction (SC-FI1 test ) , latex slide test, bentonite flocculation test, Streptococcus-L-agglutination and other procedures were carried out as well. Table 1 shows the distribution of the different methods. Without entering into details it can be stated that among the 26 Institutes participating in the collaborative assay, in this respect not less than 12 variations occurred. Similar lack of uniformity existed referring to the technologic prescriptions as specified in the protocols. With the sensitized sheep cell reaction the sensitizing doses of hemolysin varied from 1/10 to 1/2 basic agglutination titer; in some cases the titration had been performed on a hemolytic unit basis. With the latex fixation test, latex particles of difFerent origin and size as well as gamma globulins from different sources were used. The latter were either liquid or in the freeze-dried state, mostly unfiltered and not heat-treated prior to use. In some cases, however, heat treatment at 56 or 63 C., respectively, was prescribed. No identity could be stated regarding the temperature and period of incubation or the mode of reading the results. In other words, there existed almost as many modifications as study groups. This situation is self-evident when one considers that rheumatoid arthritis serology works with reagents of biological nature and therefore of divergent peculiarities. The agglutinability of red blood cells taken from different sheep or even from the same animal but at different intervals may considerably From the National Institute of Rheumatism and Mediccil IIrJdrologv, Ihdapest, Hungary. *We are indebted to Dr. K. Ujhelyi, State Institute of Hygime, Budapest, Hungary, for freeze-drying the serum samples. 891 ARTHIWITSAND RHEUMATISM, VOL.6, No. 5 ((9cToxsm),1963 642 BOZS6KY T a i l s perfomad Immbar of oaaar 5 8 1 t e r t 4 Id-II-Caat 3 S8C-taat 4 ldmx r u d e t e a t 4 SW-teat 4 bP-XI-tart SSC-teat SlC-teat + letax e l i d e taat + aC-&eat + b#-XI-taat + L-Y-XI-tart SBO-tent 4 tP-11-tent + SC-Y-XI-tent SSC-tent b bentonite floooalution taat 950-tent + L-P-XI-teat + Stroptoooooua-L-rggl. SSC-tent only 2 9 1 1 1 5 SIC-tent only 50-Y-XI-teat 2 1 only 1 e bb’-IX-trnt only rota1 a6 differ from each other,lV2 as well as many of the sensitizing capacities of individual hemolysin preparations-that is, rabbit immune sera being in reality a mix‘ture of many different kinds of materials and also containing, among others, hemolysin molecules again of different size and avidity.3 This is the reason why each laboratory has to titrate its biological reagents by trial and error when introducing the sensitized sheep cell test. Although working with far more defined substances, the situation is just as complicated with the latex fixation test. Our experience has shown that two latex suspensicns of different sources may differ in quality from each other much more markedly than any blood samples taken from individual sheep4 or supposedly even from individual alligators. In addition, F-I1 preparations expected to replace hemolysin with the promise of greater uniformity introduced a number of new problems. Namely, during the course of manufacture, gamma globulin molecules undergo changes like partial aggregation and denaturdtion, resulting in marked variations of r e a ~ t i v i t y , ~while .~ in the case cf anti-sheep rabbit sera, the disturbing effect of their several components may be reduced by simply diluting as necessary to obtain the sensitizing level. On the ground of the aforementioned arguments, the lack of uniformity ruling over the technologic prescriptions might eventually serve as a basis for the comparability of the titration results. Unfortunately, it proved not to be the case. As is seen in figure 1, the positive reference serum was found to exhibit titers between 1:32 and about 1:30,000 by the sensitized sheep cell reaction, those between 1:64 and about 1:20,000by the latex fixation test, and those between 1:64 and 1:1,280 by the STANDARDIZATION IN RHEUMATOID ARTHRITIS SEROLOGY 643 0 me0 1oMO - 0 . 51eO 6 2560 0 1880 640 . am. l60 80 . ,.. 0.0 .. 0. Ob4 0 0 0 6.b 0 4 0 . . 0.. .. . ZO - Fig. 1.-Titration results obtained by the SSC, LF-11 and SHC tests performed at different laboratories. sensitized human cell reaction. The negative serum pool was declared negative by all of the reactions performed. With the sensitized sheep cell test a value of 1:64 was also found, it is true. At the laboratory in question, however, the titer of 1:128 was taken as the borderline value of positivity in this reaction. Figure 2 shows the titration results of the positive serum pool as obtained by the sensitized sheep cell test and the latex fixation reaction performed simultaneously. Again a fair disagreement can be observed between the two tests. The titrations were performed at different points of time. According to Robinson and his collaborators7 and Swahn and Grubb,s however, freeze-drying does nct influence the agglutinating capacity of the sera. When salt-concentrated fractions of rheumatoid factor were lyophilized, all activity was lost. Nevertheless, whole plasma containing the active fraction may be kept indefinitely when treated in this manner. The data presented speak by themselves for the need of standardization in rheumatoid arthritis serology. The question remains, however, if there is a possibility of standardization. One of the main difficulties seems to be due to the well-known complexity of the so-called rheumatoid factors. LoSpalluto and Ziff,B Heimer and his collaborators,I0 Winblad,ll Harboe12 and others succeeded in demonstrating the heterogeneity of these macroglobulins. It must be referred, however, to real antibodies, where the complexity and heterogeneity of the molecules to be measured is rather a rule than an e~cepti0n.l~ In spite of this complexity, the determination of antibodies against diphtheria and tetanus toxins, strepto- 644 BOZS6KY 0 a b "i 10 Fig. %-Comparison of the titration results of the positive serum pool obtained by the SSC and LF-I1 tests performed simultaneously. lysin-0, typhoid agglutinogens, etc. may be controlled by means of proper international standard sera. Thus, differences in molecular weight and other physicochemical peculiarities do not exclude by themselves the possibility of standardization. The question is, whether there exist also some functional differences among the different rheumatoid factors. It is well-known that the rheumatoid factors can be divided into two main groups: one group reacting both in the sensitized sheep cell and latex F-I1 systems, the second one reacting only with human gamma globulin, not with rabbit gamma globulin.9The amount of these two kinds of rheumatoid factors is not proportional in individual sera. Therefore, in introducing an eventual standard serum, one has to take into account this circumstance. This problem is again not unfamiliar to those interested in the production of bacterial agglutinating sera, where antibodies of differing or shared specificity or of cross-reactivity may co-exist in the same serum. Standardization can .be solved even in instances like this, the Standard Antityphoid Serum introduced by Felix serving as an e~amp1e.l~ However, a more serious drawback in elaborating an international standard rheumatoid serum would arise, if the existence of several functionally differing rheumatoid factor groups were demonstrated. According to Kunkel and Fudenberg,15 at least eight different rheumatoid factors could be identified on the basis of reactivity with different Rh antisera. This kind of iso-specificity, however, could not be demonstrated with any of the other systems. Gamma globulin aggregates from sera of different genotypes reacted without exception 645 STANDARDIZATION IN RHEUMATOID ARTHRITIS SEROLOGY -roe. 02 -1. M44 a b44 - v I0 g r ! g o( g g 8 b g Fig. 3.-Titration of agglutinating activity of test sera in parallel with standard serum in the LF-I1 test. with all rheumatoid factor preparations obtained from different sera by ultracentrifugation. On the other hand, Butler and Vaughan,l6binding individual gamma globulin preparations with bis-diazotized benzidine bonds to human 0 erythrocytes, tested these systems with different rheumatoid euglobulin preparations. Although all of the systems employed were agglutinated in a comparable manner by a rabbit anti-human globulin serum, indicating that the degree of coating was comparable, the same cells exhibited striking differences in their ability to be agglutinated by individual rheumatoid factor preparations. This problem awaits further elucidation. The most valid argument for the possibility of establishing an international standard rheumatoid arthritis serum is based upon the existence of local reference sera, their use being recommended by Ball1' and Svartz and Schlossmann,ls early in the fifties. These authors stressed the advantages resulting from the simultaneous testing of sera of known strength by routine work. Thus, the introduction of an international standard serum, allowing calibration of the reference sera in terms of the International Unit, would supposedly meet the approval of major laboratories, especially since population studies on a large scale have been initiated lately throughout the world. The establishment of a standard serum would be of particular significance on the field of inhibition tests as well. One has to agree with Epstein,l9 that until a source of reference sera is not available, results of inhibition tests cannot be accurately compared, Differences in the level of sensitivity with the same methods employed at different laboratories or with different tests of the same laboratory would be greatly reduced if titers were replaced by potency units, as proposed by JerneZoand Miles,21 the unit of agglutinating potency being the specific activity of an arbitrarily decided weight of a dried preparation of standard serum. The potency, as the difference between logarithmic titers for the serum under test and the standard serum examined in the same experiment, 646 BOZS~KY can be calculated by dividing the reciprocal titer for the serum in question by the titer of the standard serum. Figure 3 shows the log-dose-response curves of two sera and that of a reference serum tested with the latex fixation reaction, under the conditions recommended by Singer and Plotz.22 It can be stated that the distance in degree of dilution between two tubes of the individual series showing agglutination amounted to 1280:320, i.e., 4:l in the case of serum-1, and to 80:320, i.e., 1:0.25 in the case of serum-2. If the potency of 100 units per ml. had been arbitrarily assigned to the reference serum, the potency of test serum-1 could be expressed as 4 x 100, i.e.,400 units per ml., that of test serum-2 as 0.25 x 100, i.e.,25 units per ml., irrespective of the sensitivity of the method used for titration. The unparallelism between the reaction curves for the reference serum and test serum-2 reflects differences in avidity of the rheumatoid factor molecules in question and means that in this case a generally valid relative potency does not exist. The same conclusion would be drawn if the curves were parallel but showed different distances when different agglutinating systems were used. The problem of avidity was exhaustively dealt with some years ago by I p ~ e nand ~ ~J e ~ - n ein~ ~ connection with different antibodies. Let us mention that recently differences were found also by Spaun and his collaborators25 between the slopes of log-dose-response curves of some test sera and that of the International Standard for Antistreptolysin-0, the differences being most marked in the case of low titer sera. This circumstance, however, does not invalidate the value of the antistreptolysin-0 standard serum for routine work. The data presented seem to add some practical evidence to the arguments for an international rheumatoid arthritis standard serdm. While accepting all of the difficulties that wculd be met on the way, it is encouraging to recall the period when experts had very similar problems with the standardization of serologic tests for syphilis. In this field a number of various reactions exists, the mechanisms of which are neither identical nor completely understood, a situation extremely similar to that we are facing. In spite of this, an International Standard for Human Syphilis Serum is already available.26 Similarly, the establishment of an International Standard for Rheumatoid Arthritis Serum would be one of the most important tasks for those interested in rheumatoid arthritis research work. ++ APPENOIX AUTHORS AND LABOBATORIES PARTICIPATING IN THE COLLABORATIVE ASSAY Canada D. K. Ford and H. S. Robinson Department of Medicine, University of British Columbia Vancouver, B. C. Czechoslovakia V. Houba and Fr. Lenoch V9zkumn.f Ustav Chorob Reiimatickj.ch Prague E. Kresinek and S. Sitay V9zkumnp Ustav Reumatickfrch Chor6b PieStany Finland 0. Wager Auroran Sairaala, Bakteriologinen laboratorio Helsinki STANDARDILATION IN RHEUMATOID ARTI-IBITJS SEROLOGY France A. I. Nesterov and V. I. Sachkov Y. Badin Centre d’Exploration Fonctionelle de la Seine Paris M. F. Kahn and S. de Shze H6pital LariboisiBre, Centre de Rhumatologie Viggo-Petersen Paris Germany F. Harter Medizinisch-Diagnostiisches Institct Heidelherg H. Seifert Sachsisches Serumwerk A.G. Dresden Ef ungary J. Bobory and Gy. PetrCnyi OrvostudomCnyi Egyetem, 11. Belklinika Debrecen L. Gofman and A. HAmori OrvostudomCnyi Egyetem, 11. Belklinika National Institute of Rheumatism Moscow P&s 8. Schulhof and S. Bozs6ky Orszigos Reuma 6s Furdougyi Int6zet Budapest Italy A. Robecchi Centro di Reumatologia, Ospedale Molinette Torino Xetherlunds J. J. deBlkcourt Department of Rheumatology, University Hospital Groningen W. Hijmans Department of Rheumatology, University Hospital Leiden R~mnZa I. Stoia Centrul Metodologic de Reumatologie Bucharest soviet unim 647 Sweden N. Svartz Konung Gustaf V:s forskninginstitut Stockholm S. Winblad Lunds Universitet, Institutionen fijr Klini ;k Bakteriologi hlalmij Allmanna Sjtikhus Malmii Sdtzerland E. Meyer and G . H. Fallet Centre $Etude des Maladies Rhuinatismales, Policlinique Universitaire de M6decine Geneva United Kingdom J. J . R. Duthie Rheumatic Unit, Northern General Hospital Edinburgh, Scotland United States of Arnerica J. J . Bunim National Institutes of Health, National Institute of Arthritis and Metabolic DiTeases Bethesda, Md. W. Epstein University of California Medical Center San Francisco, Calif. C. McEwen and M. Tanner Department of Medicine, New York University, Bellewe Medical Center New York, N. Y. Ch. M. PIotz Arthritis Clinic of the Mount Sinai Hospital New York, N. Y. J. H. Vaughan University of Rochester Medical Center Rochester, N. Y. M. Ziff Department of Internal Medicine, Southwestern Medical School Dallas, Texas ACKNOWLEDGMENT We are indebted to the authors for their worthy contribution and R. W. LamontHavers, Medical Director of the Arthritis and Rheumatism Foundation, New York, N. Y., for his valuable help in organizing the collaborative studies in the United States. REFERENCES 1. Zarafonetis, C. J. D., and Oster, L. H.: Heterophile agglutination variability of erythrocytes from different sheep. J . Lab. & Clin. Med. 34~1216,1946. 2. Keiper, T. W.: Pitfalls in the use of sheep cells in (?-fixation and heterophilic antibody reactions. Am. J. Clin. Path. 15:66, 1945. 648 BOZS~KY 3. Rockey, J. H., and Kunkel, H. G.: tismo, Torino, Italy. Minerva med. Studies of the rabbit antibodies which 1:298, 1961. sensitize red blood cells for agglu- 12. Harboe, M.: Heterogeneity of the tination by rheumatoid factor. Proc. “rheumatoid factor.” Atti del X. 7th Interim Scientific Session of the Congresso della Lega Internazionale American Rheumatism Association, contro il Reumatismo, Torino, Italy. December 1960, Dallas, Texas. Minerva med. 1:282, 1961. 4. Bozsbky, S.: Data on the mechanism 13. Schultze, H. E.: Immunologie und of the latex agglutination and inhiImmunochemie der Antigen-Antibition reaction. Atti del X. Congresso korperreaktionen. Zentralbl. Bakt. I. della Lega Internazionale contro il Orig. 184:324, 1962. Reumatismo, Torino, Italy. Minerva 14. Felix, A.: The provisional international med. 2:727, 1961. standard antityphoid serum. Bull. 5. Singer, J. M., Altmann, G., GoldenWorld Health Organ. 10:911, 1954. berg, A., and Plotz, Ch.M.: The 15. Kunkel, H.G., and Fudenberg, H. H.: mechanism of particulate carrier reAuto- and iso-specificity of rheumaactions with rheumatoid sera. 11. toid factors for gamma globulin. Atti Sensitizing capacity of various hudel X. Congresso della Lega Interman gamma globulins for latex nazionale contro il Reumatismo, particles. Arth. & Rheumat. 3:515, Torino, Italy. Minerva med. 1 :292, 1960. 1961. 6. Oreskes, J., Singer, J. M., and Plotz, 16. Butler, V. P., Jr., and Vaughan, J. H.: Ch.M.: Studies of human gamma Studies on the specificity of rheumaglobulin reactivity with rheumatoid toid factor for various gamma globuarthritis sera. Atti del X. Congresso lins. Atti del X. Congresso della Lega della Lega Internazionale contro il Internazionale contro il Reumatismo, Reumatismo, Torino, Italy. Minerva Torino, Italy. Minerva med. 1:293, med. 2:796, 1961. 1961. 7. Robinson, A. R., Stulberg, C. S., and 17. Ball, J.: Serum factor in rheumatoid Kuyper, A. C.: Identification of the arthritis agglutinating sensitized substance active in sheep cell agglusheep red cells. Lancet 2520, 1950. tination test for rheumatoid arthritis. 18. Svartz, N., and Schlossmann, K.: The Proc. SOC.Exper. Biol. & Med. 85:4, hemagglutination test with sensitized 1954. sheep cells in rheumatoid arthritis 8. Swahn, B., and Grubb, R.: Purification and some other diseases. Acta med. of hemagglutinating factors in rheuscandinav. 142:420, 1952. matoid arthritis sera by Cohn’s meth- 19. Epstein, V. W.:Inhibition test in the od 10 and ultracentrifugation. Acta diagnosis of rheumatoid arthritis. path. et microbiol. scandinav. 42: Arth. & Rheumat. 2347, 1959. 173, 1958. 20. Jerne, N. K.: The “unit” in preference 9. LoSpalluto, J., and Ziff, M.: Chromatoto the “titre” as a measure of aggraphic studies of the rheumatoid facglutinating activity. Bull. World tor. J. Exper. Med. 110:169, 1959. Health Organ. 10:937, 1954. 10. Heimer, R., Schwartz, E. R., and Frey- 21. Miles, A. A.: The case for unit natation berg, R. H.: Different rheumatoid in specifying the agglutinating pofactors in the serum of one patient tency of standard antisera. Bull. with rheumatoid arthritis. J. Lab. & World Health Organ. 10:941, 1954. Clin. Med. 57:16, 1961. 22. Singer, J. M., and Plotz, Ch.M.: The 11. Winblad, S.: The dependence of gamma latex fixation test. Am. J. Med. 21: globulin as reactant for demonstra888, 1956. tion of rheumatoid arthritis serum 23. Ipsen, J.: A standard for antistreptolyfactor. Atti del X. Congresso della sin-0 of human serum and its practiLega Internazionale contro il Reumacal application. Acta path. et micro- STANDARDIZATION I N RHEUMATOID ARTHRITIS SEROLOGY biol. scandinav. 21:203, 1944. 24. Jerne, N. K.: A study of avidity based on rabbit skin responses to diphtheria toxin-antitoxin mixtures. Acta path. et microbiol. scandinav. suppl. 87, 1951. 25. Spaun, J., Bentzon, M. W., Larsen, S. 649 O., and Hewitt, L. F.: International Standard for Antistreptolysin-0. Bull. World Health Organ. 24:271, 1961. 26. Bentzon, M. W., and Gag, P.: T h e International Standard for Human Syphilitic Serum. Bull. World Health Organ. 24557, 1961. S . Bozsdky, M.D., Research Assmiate, N a $ b n d Imtitute of R h m m h and Medical Hydrology, Budapest, Hungary; a$ present, Research Fellow, Rheumatic D i s m . Study ~ ~ Group, N e w Ywk University School of Medicine, B e l l m e Medical C m m , N. Y.