AMERICAN JOURNAL OF REPRODUCTIVE IMMUNOLOGY 28:226-227 © 1992 MUNKSGAARD
Remission of Rheumatoid Arthritis During Pregnancy and Maternal-Fetal Class II Alloantigen Disparity J. LEE NELSON, KATHLEEN A. HUGHES, ANAJANE G. SMITH, BRENDA B. NISPEROS, ANN M. BRANCHAUD, AND JOHN A. HANSEN Fred Hutchinson Cancer Research Center, Division ofClinical Research, Seattle, Washington ABSTRACT: Rheumatoid arthritis (RA) is an autoimmune disorder known to be associated with specific class II genes. Although it has been known since 1938 that the majority of women with RA experience disease improvement or remission during pregnancy, the reasons remain unknown. Pregnancy represents an immunologic challenge and maternal immune recognition of the semi-allogeneic fetus occurs as part of normal pregnancy. We hypothesized that maternal immune response to fetal HLA antigens might be associated with the effect ofpregnancy on arthritis activity. To test this hypothesis, we studied HLA antigens in mother-child pairs comparing maternal-fetal HLA antigen sharing for pregnancies where arthritis improved with those where disease wasactive. No significant difference was observed in the two groups for class I HLA antigens. Fetal-maternal disparity for HLA-DR and HLA-DQ antigens was observed significantly more frequently in pregnancies with remission or improvement compared with those in which disease was active. These observations suggest that maternal immune response to fetal paternally-inherited class II HLA antigens may be important in RA remission observed during pregnancy. (Am J Reprod Immunol. 1992; 28:226-227.) Key words: Rheumatoid arthr-itis, alloantigen, disease remission, HLA-DR, HLA-DQ INTRODUCTION
In 1938 Hench first described the subsidence ofrheumatoid arthritis (RA) during pregnancy.' His research into the cause of this improvement contributed to the eventual discovery of cortisone. Subsequent work showed that elevation of cortisol levels was not the reason for improvement ofRA during pregnancy.f-" Other investigators have proposed a variety of different serum factors and pregnancy products to account for the amelioration of RA with pregnancy; however, none has withstood subsequent challenge and the reasons for disease remission remain unknown.i''" METHODS
Women meeting the 1987 revised criteria for rheumatoid arthritis" and with one or more pregnancies after RA onset were studied. Cases were referred by practicing rheumatologists and obstetricians in King and Whatcom counties over a 5-year period from 1986 through 1991. Diagnostic characteristics of the patients included the
Submitted for publication July 17, 1992; accepted July 22, 1992. Address reprint requests to J. Lee Nelson, Fred Hutchinson Cancer Research, Division of Clinical Research Center, 1124 Columbia St., Seattle, WA98104.
following: history of morning stiffness of more than one hour's duration (100%); simultaneous swelling in three or more joint areas (97%); swelling in at least one metacarpal-phalangeal, proximal interphalangel, or wristjoint (100%); symmetrical joint swelling (100%); rheumatoid nodules (31%); positive rheumatoid factor (76%); and typical erosive changes of rheumatoid arthritis on radiographic examination (87%). (Rhematoid factor positivity may be underestimated since no record of rheumatoid factor determination was available for several women who tested negatively at study entry and in whom arthritis had been severe and destructive but was inactive at study entry.) Fifty-five pregnancies of 39 women were studied. Approximately one-quarter of cases were followed prospectively and the remainder were reviewed to record physician assessment ofRA disease activity before, during, and after pregnancy. The patient's symptoms, use of arthritis medication, and the physician's evaluation and physical exam were considered in the assessment of disease activity. Patients were given a month-by-month calendar and asked to chronicle the activity of their arthritis on a scale of 0 to 3, 0 = inactive, 1 = mild activity, 2 = moderate, and 3 = severe; the calendar extended for 6 months prior to conception, throughout pregnancy, and for 3 months after delivery. Disease activity was classified as remission, improvement, inactive, or active. Criteria fur remission were that the patient have discontinued all arthritis medication, have < 15 minutes of morning stiffness, and have no joint pain or tenderness and no joint swelling on physical examination. The erythrocyte sedimentation rate rises during the course of normal pregnancy, so it was not included as a remission criterion. Criteria for improvement included a reduction or discontinuation of arthritis medication while symptoms remained better than prior to pregnancy, less than 15 minutes of morning stiffness, and no evidence of joint swelling on exam. If either corticosteroids or nonsteroidal anti-inflammatory drugs were used in the second and third trimester, a patient was classified as improved. Some pregnancies were classified as inactive, since, although symptoms ami signs of arthritis were absent during these pregnancies, a baseline assessment of disease activity in the 6 months prior to pregnancy was lacking. Therefore, some of these may represent persistent inactivity rather than remission or improvement. Women with active disease had both subjective and objective (swelling on physical examination) evidence of disease activity at a level greater than that of the 6 months prior to pregnancy and included one in which the onset of disease occurred during pregnancy. Disease-modifying medications used prior to pregnancy were discontinued during pregnancy in all cases. Once in remission (or improved), most patients sustained the effect throughout pregnancy. HLA typing was completed using a standard NIH twostage complement-dependent microcytotoxicity assay,"
CLASS II GENES AND RA REMISSION
HLA-A, HLA-B, HLA-C, HLA-DR, and HLA-DQ antigens were examined for mothers and children. For cases followed prospectively, typing for the newborns was done from lymphocytes isolated from cord blood at the time of delivery. RESULTS
No difference was observed between pregnancies characterized by full remission and those with partial remission, so these groups were combined. Comparisons were made between this group and those in which disease was active, since pregnancies in which pre-pregnancy arthritis activity was uncertain may have included some with persistence of inactive disease rather than arthritis improvement. Fetal-maternal disparities for HLA-A, HLA-B, and HLA-C antigens were not significantly different in the two groups. Fetal disparities from maternal class II alloantigens, however,were observed significantly more frequently in remitting/improving pregnancies when compared with those in which arthritis was active. This observation was true both for HLA-DR and for HLA-DQ, but was more marked for HLA-DQ. A comparison of clinical characteristics showed no difference between the two groups that might alternatively explain this finding. There was no significant difference between the groups in rheumatoid factor status, maternal age at RA onset, or at study pregnancy, or birth order of study pregnancy. The RA susceptibility antigen HLADR4 was not different for women with pregnancies in the improving/remitting group compared with the active group. Additionally, fetal paternally-inherited HLA-DR and HLA-DQ antigens were diverse and did not differ in distribution between the two groups. DISCUSSION
Specific HLA class II antigens are associated with a wide variety of different autoimmune diseases." However, the mechanism by which particular alleles confer disease susceptibility remains largely unknown. Whether the fetal-maternal class II alloantigen disparity associated with arthritis improvement during pregnancy described in these studies is related to class II antigen participation in disease pathogenesis is unknown, and the mechanism by which a disparity might effect arthritis activity is speculative. Studies in animal models of autoimmune disease have demonstrated disease abrogation or diminution with the administration of antibodies to class II antigens.P'"! In these models, antibodies were directed to self Ia antigens; however, investigators
227
in France have reported arthritis improvement when RA patients were given placental-eluted gamma globulins'v'" containing alloantibodies to HLA class II antigens. 14 CONCLUSION
In conclusion, the pregnancy induced amelioration of rheumatoid arthritis is associated with fetal-maternal disparity for class II alloantigens. Further understanding of this association should generate insight into disease pathogenesis and hopefully proivde a background for developing new treatment strategies for rheumatoid arthritis. REFERENCES 1. Hench PS: The ameliorating effect of pregnancy on chronic atrophic (infectious rheumatoid) arthritis, fibrositis and intermittent hydrathrosis. Proc Mayo Clinic. 1938; 13:161-166. 2. Smith WD, West HF: Pregnancy and rheumatoid arthritis. Acta Rheumatol Scand. 1960; 6:189-201. 3. Persellin RH: The effect of pregnancy on rheumatoid arthritis. Bull Rheum Dis. 1976; 5:922-926. 4. Ostensen M, von Schoultz B, Husby G: Comparison between a-pregnancy associated globulin and activity of rheumatoid arthritis and ankylosing spondylitis during pregnancy. Scand J Rheumatol. 1983; 12:315-318. 5. Arnett FC, Edworthy SM, Bloch D, McShane DJ, Fries JF, Cooper NS, Healey LA, Kaplan SR, Liang MH, Luthra HS. The American Rheumatism Association revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988; 31:315-324. 6. Ray JG: In: NIAID Manual of Tissue Typing Techniques. 1979-1980. NIH Publication No. 80-545. Bethesda, 1979 pp. 39-41. 7. Nelson JL, Hansen JA: Autoimmune disease and HLA. In: Atassi MZ, ed. CRC Reviews in Immunology. Boca Raton, Ann Arbor, Boston:CRC Press, 1990:307-328. 8. Steinman L, Rosenbaum J, Sriram S, McDevitt HO: In vivo effects of antibodies to immune response gene products: Prevention of experimental allergic encephalitis. Proc Natl Acad Sci USA. 1981; 78:7111-7114. 9. Waldor M, Sriram S, McDevitt HO, Steinman L: In vivo therapy with monoclonal anti-I-A-antibody suppressed immune responses to acetylcholine receptor. Proc Nat! Acad Sci USA. 1983; 80: 2713-2717. 10. Adelman NE, Watling DL, McDevitt HO: Treatment of (NZB NZW) F1 disease with anti-I-A monoclonal antibodies. J Exp Med. 1983; 158:1350-1355. 11. Wooley PM, Lutra HS, Liabise WP, Stuart J, David CWS: Type II collagen induced arthritis in mice. III. Suppression of arthritis by using monoclonal and polyclonal anti-Ia antisera. J Immunol. 1983; 134:2361-2374. 12. Sany J, Clot J, Bonneau M, Andary M: Immunomodulating effect of human placenta-eluted gamma globulins in rheumatoid arthritis. Arthritis Rheum. 1982; 25:17-24. 13. Combe B, Cosso B, Clot J, Bonneau M, Sany J: Human placentaeluted gamma globulins in immunomodulating treatment ofrheumatoid arthritis. Am J Med. 1985; 78:920-928. 14. Moynier M, Cosso B, Brochier J, Clot J: Identification of class II HLA alloantibodies in placenta-eluted gamma globulins used for treating rheumatoid arthritis. Arthritis Rheum. 1987; 30:375-381.
Remission of Rheumatoid Arthritis During Pregnancy and Maternal-Fetal Class II Alloantigen Disparity J. LEE NELSON, KATHLEEN A. HUGHES, ANAJANE G. SMITH, BRENDA B. NISPEROS, ANN M. BRANCHAUD, AND JOHN A. HANSEN Fred Hutchinson Cancer Research Center, Division ofClinical Research, Seattle, Washington ABSTRACT: Rheumatoid arthritis (RA) is an autoimmune disorder known to be associated with specific class II genes. Although it has been known since 1938 that the majority of women with RA experience disease improvement or remission during pregnancy, the reasons remain unknown. Pregnancy represents an immunologic challenge and maternal immune recognition of the semi-allogeneic fetus occurs as part of normal pregnancy. We hypothesized that maternal immune response to fetal HLA antigens might be associated with the effect ofpregnancy on arthritis activity. To test this hypothesis, we studied HLA antigens in mother-child pairs comparing maternal-fetal HLA antigen sharing for pregnancies where arthritis improved with those where disease wasactive. No significant difference was observed in the two groups for class I HLA antigens. Fetal-maternal disparity for HLA-DR and HLA-DQ antigens was observed significantly more frequently in pregnancies with remission or improvement compared with those in which disease was active. These observations suggest that maternal immune response to fetal paternally-inherited class II HLA antigens may be important in RA remission observed during pregnancy. (Am J Reprod Immunol. 1992; 28:226-227.) Key words: Rheumatoid arthr-itis, alloantigen, disease remission, HLA-DR, HLA-DQ INTRODUCTION
In 1938 Hench first described the subsidence ofrheumatoid arthritis (RA) during pregnancy.' His research into the cause of this improvement contributed to the eventual discovery of cortisone. Subsequent work showed that elevation of cortisol levels was not the reason for improvement ofRA during pregnancy.f-" Other investigators have proposed a variety of different serum factors and pregnancy products to account for the amelioration of RA with pregnancy; however, none has withstood subsequent challenge and the reasons for disease remission remain unknown.i''" METHODS
Women meeting the 1987 revised criteria for rheumatoid arthritis" and with one or more pregnancies after RA onset were studied. Cases were referred by practicing rheumatologists and obstetricians in King and Whatcom counties over a 5-year period from 1986 through 1991. Diagnostic characteristics of the patients included the
Submitted for publication July 17, 1992; accepted July 22, 1992. Address reprint requests to J. Lee Nelson, Fred Hutchinson Cancer Research, Division of Clinical Research Center, 1124 Columbia St., Seattle, WA98104.
following: history of morning stiffness of more than one hour's duration (100%); simultaneous swelling in three or more joint areas (97%); swelling in at least one metacarpal-phalangeal, proximal interphalangel, or wristjoint (100%); symmetrical joint swelling (100%); rheumatoid nodules (31%); positive rheumatoid factor (76%); and typical erosive changes of rheumatoid arthritis on radiographic examination (87%). (Rhematoid factor positivity may be underestimated since no record of rheumatoid factor determination was available for several women who tested negatively at study entry and in whom arthritis had been severe and destructive but was inactive at study entry.) Fifty-five pregnancies of 39 women were studied. Approximately one-quarter of cases were followed prospectively and the remainder were reviewed to record physician assessment ofRA disease activity before, during, and after pregnancy. The patient's symptoms, use of arthritis medication, and the physician's evaluation and physical exam were considered in the assessment of disease activity. Patients were given a month-by-month calendar and asked to chronicle the activity of their arthritis on a scale of 0 to 3, 0 = inactive, 1 = mild activity, 2 = moderate, and 3 = severe; the calendar extended for 6 months prior to conception, throughout pregnancy, and for 3 months after delivery. Disease activity was classified as remission, improvement, inactive, or active. Criteria fur remission were that the patient have discontinued all arthritis medication, have < 15 minutes of morning stiffness, and have no joint pain or tenderness and no joint swelling on physical examination. The erythrocyte sedimentation rate rises during the course of normal pregnancy, so it was not included as a remission criterion. Criteria for improvement included a reduction or discontinuation of arthritis medication while symptoms remained better than prior to pregnancy, less than 15 minutes of morning stiffness, and no evidence of joint swelling on exam. If either corticosteroids or nonsteroidal anti-inflammatory drugs were used in the second and third trimester, a patient was classified as improved. Some pregnancies were classified as inactive, since, although symptoms ami signs of arthritis were absent during these pregnancies, a baseline assessment of disease activity in the 6 months prior to pregnancy was lacking. Therefore, some of these may represent persistent inactivity rather than remission or improvement. Women with active disease had both subjective and objective (swelling on physical examination) evidence of disease activity at a level greater than that of the 6 months prior to pregnancy and included one in which the onset of disease occurred during pregnancy. Disease-modifying medications used prior to pregnancy were discontinued during pregnancy in all cases. Once in remission (or improved), most patients sustained the effect throughout pregnancy. HLA typing was completed using a standard NIH twostage complement-dependent microcytotoxicity assay,"
CLASS II GENES AND RA REMISSION
HLA-A, HLA-B, HLA-C, HLA-DR, and HLA-DQ antigens were examined for mothers and children. For cases followed prospectively, typing for the newborns was done from lymphocytes isolated from cord blood at the time of delivery. RESULTS
No difference was observed between pregnancies characterized by full remission and those with partial remission, so these groups were combined. Comparisons were made between this group and those in which disease was active, since pregnancies in which pre-pregnancy arthritis activity was uncertain may have included some with persistence of inactive disease rather than arthritis improvement. Fetal-maternal disparities for HLA-A, HLA-B, and HLA-C antigens were not significantly different in the two groups. Fetal disparities from maternal class II alloantigens, however,were observed significantly more frequently in remitting/improving pregnancies when compared with those in which arthritis was active. This observation was true both for HLA-DR and for HLA-DQ, but was more marked for HLA-DQ. A comparison of clinical characteristics showed no difference between the two groups that might alternatively explain this finding. There was no significant difference between the groups in rheumatoid factor status, maternal age at RA onset, or at study pregnancy, or birth order of study pregnancy. The RA susceptibility antigen HLADR4 was not different for women with pregnancies in the improving/remitting group compared with the active group. Additionally, fetal paternally-inherited HLA-DR and HLA-DQ antigens were diverse and did not differ in distribution between the two groups. DISCUSSION
Specific HLA class II antigens are associated with a wide variety of different autoimmune diseases." However, the mechanism by which particular alleles confer disease susceptibility remains largely unknown. Whether the fetal-maternal class II alloantigen disparity associated with arthritis improvement during pregnancy described in these studies is related to class II antigen participation in disease pathogenesis is unknown, and the mechanism by which a disparity might effect arthritis activity is speculative. Studies in animal models of autoimmune disease have demonstrated disease abrogation or diminution with the administration of antibodies to class II antigens.P'"! In these models, antibodies were directed to self Ia antigens; however, investigators
227
in France have reported arthritis improvement when RA patients were given placental-eluted gamma globulins'v'" containing alloantibodies to HLA class II antigens. 14 CONCLUSION
In conclusion, the pregnancy induced amelioration of rheumatoid arthritis is associated with fetal-maternal disparity for class II alloantigens. Further understanding of this association should generate insight into disease pathogenesis and hopefully proivde a background for developing new treatment strategies for rheumatoid arthritis. REFERENCES 1. Hench PS: The ameliorating effect of pregnancy on chronic atrophic (infectious rheumatoid) arthritis, fibrositis and intermittent hydrathrosis. Proc Mayo Clinic. 1938; 13:161-166. 2. Smith WD, West HF: Pregnancy and rheumatoid arthritis. Acta Rheumatol Scand. 1960; 6:189-201. 3. Persellin RH: The effect of pregnancy on rheumatoid arthritis. Bull Rheum Dis. 1976; 5:922-926. 4. Ostensen M, von Schoultz B, Husby G: Comparison between a-pregnancy associated globulin and activity of rheumatoid arthritis and ankylosing spondylitis during pregnancy. Scand J Rheumatol. 1983; 12:315-318. 5. Arnett FC, Edworthy SM, Bloch D, McShane DJ, Fries JF, Cooper NS, Healey LA, Kaplan SR, Liang MH, Luthra HS. The American Rheumatism Association revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988; 31:315-324. 6. Ray JG: In: NIAID Manual of Tissue Typing Techniques. 1979-1980. NIH Publication No. 80-545. Bethesda, 1979 pp. 39-41. 7. Nelson JL, Hansen JA: Autoimmune disease and HLA. In: Atassi MZ, ed. CRC Reviews in Immunology. Boca Raton, Ann Arbor, Boston:CRC Press, 1990:307-328. 8. Steinman L, Rosenbaum J, Sriram S, McDevitt HO: In vivo effects of antibodies to immune response gene products: Prevention of experimental allergic encephalitis. Proc Natl Acad Sci USA. 1981; 78:7111-7114. 9. Waldor M, Sriram S, McDevitt HO, Steinman L: In vivo therapy with monoclonal anti-I-A-antibody suppressed immune responses to acetylcholine receptor. Proc Nat! Acad Sci USA. 1983; 80: 2713-2717. 10. Adelman NE, Watling DL, McDevitt HO: Treatment of (NZB NZW) F1 disease with anti-I-A monoclonal antibodies. J Exp Med. 1983; 158:1350-1355. 11. Wooley PM, Lutra HS, Liabise WP, Stuart J, David CWS: Type II collagen induced arthritis in mice. III. Suppression of arthritis by using monoclonal and polyclonal anti-Ia antisera. J Immunol. 1983; 134:2361-2374. 12. Sany J, Clot J, Bonneau M, Andary M: Immunomodulating effect of human placenta-eluted gamma globulins in rheumatoid arthritis. Arthritis Rheum. 1982; 25:17-24. 13. Combe B, Cosso B, Clot J, Bonneau M, Sany J: Human placentaeluted gamma globulins in immunomodulating treatment ofrheumatoid arthritis. Am J Med. 1985; 78:920-928. 14. Moynier M, Cosso B, Brochier J, Clot J: Identification of class II HLA alloantibodies in placenta-eluted gamma globulins used for treating rheumatoid arthritis. Arthritis Rheum. 1987; 30:375-381.
