- Email: [email protected]
The presence of serum antibody to the chlamydial heat shock protein (CHSP60) as a diagnostic test for tubal factor infertility
FERTILITY AND STERILITY ®
Vol. 67, No. 3, March 1997
Copyright ' 1997 American Snciety for Reproductive Medicine
Printed on acid-free paper in U. S. A.
The presence of serum antibody to the chlamydial heat shock protein (CHSP60) as a diagnostic test for tubal factor infertility*
Paul Claman, M.D.t$ Lisa Honey, M.D.t Rosanna W. Peeling, Ph.D.§
Peter Jessamine, M.D.[[ Baldwin Toye, M.D.¶
Ottawa Civic and General Hospitals, University of Ottawa, Ottawa, Ontario, and the Laboratory Centre for Disease Control, Winnipeg Manitoba, Canada
Objective: To study the utility of testing for heat shock protein 60 (CHSP60) antibodies in the diagnosis of tubal factor infertility. Design: Prospective case control. Setting: Canadian university hospital infertility clinic. P a t i e n t ( s ) : Women presenting for infertility investigation. I n t e r v e n t i o n ( s ) : Sera were collected from 77 patients. Main O u t c o m e Measure(s): The relationship between tubal factor infertility and the presence of antibodies to Chlamydia trachomatis and CHSP60 was assessed. Result(s): There were no significant differences between antibodies to C. trachomatis in women with tubal factor infertility (63%) and other causes of infertility (46%). However, more women with tubal factor infertility (44%) had anti-CHSP60 antibodies compared with other causes of infertility (8%). Antibody testing for C. trachomatis has only a 63% sensitivity and a 54% specificity for detecting tubal factor infertility. In contrast, the CHSP60 antibody test has a 44% sensitivity and a 92% specificity for detecting tubal factor infertility. There is a good positive likelihood ratio of 5.5 for CHSP60 antibody testing detecting the presence of tubal factor infertility. Combining CHSP60 antibody with antibody testing for C. trachomatis has an excellent positive likelihood ratio of 10 for the detection of C. trachomatis-associated tubal factor infertility. Conclusion(s): CHSP60 antibody testing is a more accurate test than antibody testing for C. trachomatis for predicting chlamydia-associated tubal factor infertility. These tests, when used in combination at initial infertility evaluation, would provide a rapid noninterventive means of diagnosing tubal factor infertility. Fertil Steril e 1997;67:501-4 Key Words: Chlamydia heat shock protein 60 (CHSP60), Chlamydia trachomatis, tubal infertility, infertility
Infection with C h l a m y d i a t r a c h o m a t i s is an imp o r t a n t sexually t r a n s m i t t e d disease t h a t h a s been associated with t u b a l infertility in women. The risk
Received July 25, 1996; revised and accepted October 28, 1996. * Supported in part with a grant from Physician Services Incorporated Foundation, and National Laboratory for Sexually Transmitted Diseases, Laboratory Centre for Disease Control, Ottawa, Ontario, Canada. t Division of Reproductive Medicine, Department of Obstetrics and Gynecology,Ottawa Civic Hospital, University of Ottawa. $ Reprint requests: Paul Claman, M.D., 552-737 Parkdale Avenue, Ottawa, Ontario, Canada K1Y4E9 (FAX: 613-724-4942; email: [email protected]). § Laboratory Centre for Disease Control. Vol. 67, No. 3, March 1997
of t u b a l infertility a p p e a r s to i n c r e a s e with the n u m ber a n d severity of episodes of salpingitis r e s u l t i n g in t u b a l occlusion (1). M a n y studies h a v e s h o w n t h a t serologic evidence of prior C h l a r n y d i a infection is s t r o n g l y associated with t u b a l infertility ( 1 - 4 ) . The m e c h a n i s m by w h i c h c h l a m y d i a l infection results in t u b a l d a m a g e h a s been s t u d i e d widely b u t r e m a i n s unclear. P r i m a t e studies s u g g e s t t h a t rein[[Division of Microbiologyand Immunology, Department of Laboratory Medicine, Division of Infectious Disease, Department of Medicine, Ottawa Civic Hospital, University of Ottawa. ¶ Division of Microbiology, Department of Pathology and Laboratory Medicine, Division of Infectious Disease, Department of Medicine, Ottawa General Hospital, University of Ottawa. Claman et al. CHSP60 as a test for tubal infertility
501
fection with C. t r a c h o m a t i s may be the necessary component for chronic salpingitis leading to distal tubal obstruction (5). The immunopathology seen in genital tract infections is similar to those elicited by the chlamydial conjunctival infection leading to scarring trachoma. Reinfection has been identified as an important risk factor in the pathogenesis of trachoma (6). Recent studies have shown a strong association between antibody response to the chlamydia heat shock protein 60 (CHSP60) and ectopic pregnancy (7) as well as the development of chla] mydia-associated tubal infertility (8). Chlamydial heat shock protein is a homolog of the gro-EL family of heat shock proteins (9, 10). This family of proteins is highly conserved among both prokaryotes and eukaryotes (11, 12). It is been suggested that antibodies against conserved epitopes on CHSP60 may cross react with those of h u m a n HSP60 and initiate an autoimmune inflammatory response (7, 13-16). A recent study identified testing for serum antibodies to the whole C. t r a c h o m a t i s organism as more accurate than hysterosalpingography (HSG) in predicting tubal factor infertility (4). We compared the accuracy of chlamydia serology and antibodies against CHSP60 in predicting the diagnosis of tubal factor infertility in a prospective study. MATERIALS AND METHODS
After the proposed study and consent forms were approved by the Research Ethics Committee at the Ottawa Civic Hospital, sera were obtained from 77 unselected women presenting for initial infertility evaluation in the infertility clinic at the University of Ottawa, Ottawa Civic Hospital. All women had standard investigations performed that included BBT charting, and/or midluteal P, and/or late luteal endometrial biopsy, cervical C. t r a c h o m a t i s screening with ELISA or polymerase chain reaction assays. The diagnosis of tubal infertility was made by HSG and/or laparoscopic examination demonstrating distal tubal obstruction or laparoscopic evidence of severe peritubal adhesions. If complete bilateral distal tubal obstruction was found on HSG, a laparoscopic examination was not always performed. All sera were assayed for immunoglobulin (Ig) G and IgM antibody to C. t r a c h o m a t i s by the microimmunofluorescence (MIF) technique of Wang and Grayston (17) using purified Formalin-fixed elementary bodies. Sera were screened at a dilution of 1:8 and titered at twofold dilutions to end point. C h l a m y d i a t r a c h o m a t i s seropositivity was defined as an MIF titer of - 1 : 8 (18). An ELISA using recombinant CHSP60 expressed as a fusion protein with glutathione-S-transferase 502
Claman et al.
CHSP60 as a test for tubal infertility
as an antigen was used to test for the presence of antibody to CHSP60 as described previously (8, 19). Patient sera were diluted 1:500 and incubated with recombinant antigen bound to 96-well microtiter plates. Horseradish peroxidase-conjugated goat anti-human IgG was added and the optical density o f each well was determined. All sera positive by ELISA were confirmed by immunoblotting using recombinant CHSP60 as antigen. Sera were processed blindly and without knowledge of the clinical diagnosis of the patient. Groups were compared with X2 or Fisher's exact test. Odds ratios (OR) with 95% confidence intervals (CI) also were calculated. Likelihood ratios also were computed to help evaluate the diagnostic accuracy of both serologic tests for their ability to predict the presence of tubal disease as the cause of infertility in the patient presenting for infertility evaluation. The calculation of likelihood ratios enables comparison of the diagnostic values of tests independent of prevalence of infection in differing populations. The likelihood ratio of a positive test is calculated as [sensitivity/(100 - specificity)]. A positive likelihood ratio of 2 to 5 indicates a fair clinical test, 5 to 10 indicates a good clinical test, and > 10 is an excellent clinical test. The likelihood ratio of a negative test is calculated as [(100 - sensitivity)/specificity]. A negative likelihood ratio of 0.5 to 0.2 indicates a fair clinical test, 0.2 to 0.1 is a good clinical test, and <0.1 is an excellent clinical test (20). Because likelihood ratios refer to the actual test results before the disease status is known, they are more useful to clinicians than are sensitivity and specificity (21). RESULTS
Seventy-seven women were studied, 16 with the final diagnosis of tubal disease. Two of those with tubal disease also had ovulatory dysfunction, three also had male factor problems and, three had both male factor and ovulatory dysfunction in addition to tubal disease. Sixty-one patients had nontubal causes of infertility: 13 with male factor, 14 with ovulatory dysfunction, with 17 mixed male factor and ovulatory dysfunction, and 17 with idiopathic infertility. There was no difference between the mean age of the women with tubal infertility (32.6 _ 3.2 years) and those with other causes of infertility (31 _+ 4.9 years). A total of six women had a history of pelvic inflammatory disease (PID), but only three of these were found to have tubal disease. Thus, 13 of 16 women with tubal disease in our series had no prior history of PID. Thirteen patients were found to have endometriosis at the time of laparoscopy, and nine of these women had no distortion of tuboovarian anatomy. Seven patients had a prior history Fertility and Sterility ®
of ectopic pregnancy, and three of these women appeared to have normal tubes at laparoscopy after resolution of the ectopic pregnancy (treated with methotrexate or linear salpingostomy). The rest of our patients had a final diagnosis of either male or ovulatory factor infertility. A positive MIF titer was found in 10 of 16 (63%) women with tubal disease and in 28 of 61 (46%) with other causes of infertility (P = 0.2, OR = 1.96, 95% CI = 0.56 to 7.1). Thus, as a screening test for tubal disease, MIF has a sensitivity of 63% and a specificity of 54%, with a positive predictive value of 26% and a negative predictive value of 85%. Calculation of likelihood ratios also show t h a t MIF testing for C. trachomatis antibodies is not a useful clinical test for the prediction of tubal disease in the patient presenting with infertility (positive likelihood ratio = 1.36 and negative likelihood ratio = 0.69). The CHSP60 antibodies were found in 7 of 16 (44%) women with tubal disease and in only 5 of 61 (8%) women with other causes of infertility (P = 0.002, OR = 8.7, 95% CI = 1.9 to 41.9, power = 0.976 for c~ = 0.05). Therefore, as a m a r k e r of tubal infertility, CHSP60 antibody testing has a sensitivity of 44% and a specificity of 92%. For the detection of C. trachomatis-associated tubal disease (defined as tubal disease with a positive MIF titer), the CHSP60 test was positive in 7 of 10 patients with tubal disease and in only 5 of 67 patients with other causes of infertility or with a negative MIF titer (P = 0.0000325, OR = 29, 95%CI = 5 to 215). Thus, the sensitivity of the CHSP60 assay for detecting C. trachomatis associated tubal disease is 70% and its specificity is 93%. The positive predictive value for the CHSP60 test for the diagnosis of tubal infertility is 58% and the negative predictive value is 86%. A patient with tubal disease is 5.5 times more likely than a patient without tubal disease to have antibodies to CHSP60 (positive likelihood ratio). The likelihood ratio of a negative CHSP60 test was 0.61. When both the MIF and CHSP60 test are used, the CHSP60 has a positive likelihood ratio of 10 and a negative likelihood ratio of 0.32, making it an excellent test for the detection of C. trachomatis-associated tubal disease. DISCUSSION
It has been suggested t h a t C. trachomatis antibody testing might be useful as a screening tool in predicting tubal factor infertility (22). As a predictor of tubal disease, we found t h a t the MIF for C. trachomatis test has a very limited specificity of only 54% and a sensitivity of 63% for predicting tubal disease in our clinic population. Positive and negative likelihood ratios of 1.36 and 0.69 show t h a t the MIF is a Vol. 67, No. 3, March 1997
poor clinical test for the detection of tubal disease. Dabekausen et al. (4) have suggested t h a t C. trachom a t i s antibody may be better at diagnosing tubal disease than hysterosalpingography. The antibody assay system used in their study m a y have limited specificity to differentiate between antibodies to C. trachomatis and C. p n e u m o n i a e . The seroprevalence of C. p n e u m o n i a e in most adult populations is 40% to 60% (23). The MIF test allows more specific detection of antibodies against different chlamydial species (24). Furthermore, Dabekausen et al.'s (4) definition of tubal disease included tuboperitoneal adhesions and even unilateral tubal pathology seen at laparoscopy, whereas in our series we defined tubal disease as the presence of bilateral distal tubal obstruction or severe peritubal adhesions. In this prospective study, antibody to CHSP60 was found in 44% of women with tubal factor infertility and only 8% of women with other causes of infertility. Antibody to CHSP60 was found to be a very specific predictor of the diagnosis of chlamydia-associated tubal factor infertility. Although the sensitivity of this test for detecting tubal disease is poor, the 92% specificity of this test for detecting tubal disease makes it an important assay. The positive likelihood ratio of CHSP60 antibody testing for detecting the presence of tubal disease in an unselected population of patients presenting for infertility was 5.5, making it a good clinical test for predicting the presence of tubal factor infertility. These data in this prospective study are consistent with our findings from a retrospective study of CHSP60 antibody response in patients undergoing IVF-ET t r e a t m e n t (8). The sensitivity of any test for the diagnosis of C. trachomatis-associated tubal infertility is dependent on the prevalence of prior genital chlamydial infection in the population being tested. Thus, the CHSP60 antibody test may be limited by the fact that there were patients with other causes of tubal infertility, including previous gonococcus infection, severe endometriosis, tuberculosis, and history of ectopic pregnancies. The CHSP60 antibody testing will detect only tubal damage resulting from C. trachomatis infections, which will not account for tubal disease due to other causes. The sensitivity of the CHSP60 test for detecting tubal disease in C. trachomatis MIF seropositive patients was 70% and the specificity was 93% in this study. If both the MIF and CHSP60 tests are used, the CHSP60 will have a positive likelihood ratio of 10 and a negative likelihood ratio of 0.32, making it an excellent test for .the detection of C. trachomatis-associated tubal infertility (defined as tubal infertility with a positive MIF titer). B r u n h a m et al. (7) previously reported t h a t 19 of 21 (91%) patients with ectopic pregnancy and seropositive for C. trachomatis had serum antibodies to Claman et al. CHSP60 as a test for tubal infertility
503
CHSP60. Seven women in our population sample had a prior history of ectopic pregnancy, four of whom were MIF positive. Three of these four women had CHSP60 antibodies (75%). Only four of these seven women had obvious tubal disease. One of the important reasons for testing for chlamydia as opposed to other sexually transmitted diseases as a cause of tubal disease and infertility is that the majority of genital chlamydia infections in women are asymptomatic. Undetected and untreated these infections can ascend into the upper genital tract and cause PID with resultant ectopic pregnancy arid tubal occlusion. Many studies have shown that most women with tubal infertility did not recall a history of chlamydia infection (3). These tests would be a rapid means of identifying these women and of establishing the diagnosis without other unnecessary and costly interventive procedm'es. The usefulness of the CHSP60 antibody test lies in its high specificity (92%) and good positive likelihood ratio (5.5) for the presence of tubal factor infertility. A negative MIF test has a very good (85%) negative predictive value, which makes it useful alone in ruling out tubal factor as the cause of infertility. A positive MIF assay combined with CHSP60 antibody testing makes for an excellent test ofC. trachomatisassociated tubal infertility (positive likelihood ratio = 10). The CHSP60 antibody test combined with MIF testing should be introduced as a diagnostic tool as part of the initial infertility evaluation. Patients with a positive CHSP60 antibody assay may even consider moving into IVF-ET treatment rather than pursue further interventive and expensive diagnostic testing, especially because IVF-ET appears to be particularly effective in the treatment of tubal factor infertility patients who have antibodies to the CHSP60 (25). REFERENCES 1. Westrom L. Chlamydia and its effect on reproduction. J Br Fertil Soc 1998; 1:23-30. 2. Brunham RC, MacLean IW, Binns B, Peeling RW. Chlumydia trachomatis: its role in tubal infertility. J Infect Dis 1985; 152:1275-82. 3. Cates W Jr, Wasserheit JN. Genital chlamydial infections: epidemiology and reproductive sequelae. Am J Obstet Gynecol 1991; 164:1771-81. 4. Dabekausen YAJM, Evers JLH, Land JA, Stals FS. Chlamydia trachomatis antibody testing is more accurate than hysterosalloingography in predicting tubal factor infertility. Fertil Steril 1994;61:833-7. 5. Patton DL, Kuo CC, Wang SP, Halbert SA. Distal tubal obstruction induced by repeated Chlamydia trachomatis salpingeal infections in pig-tailed macaques. J Infect Dis 1987; 155:1292-9. 6. Grayson JT, Wang SP, Yeh hi, Kuo CC. Importance of reinfection in the pathogenesis of trachoma. Rev Infect Dis 1985;7:717-25.
504
Claman et al.
CHSP60 as a test for tubal infertility
7. Brunham RC, Peeling R, MacLean I, Kosseim ML, Paraskevas M. Chlamydia trachomatis-associated ectopic pregnancy: serologic and histologic correlates. J Infect Dis 1992; 165:1076-81. 8. Toye B, Laferriere C, Claman P, Jessamine P, Peeling R. Association between antibody to the chlamydial heat-shock protein and tubal infertility. J Infect Dis 1993; 168:1236-40. 9. Morrison RP, Su H, Lyng K, Yran Y. The Chlamydia trachomatis Hyp operon is homologous to the groE stress response operon of Esherichia coli. Infect Immun 1990;58:2701-5. 10. Cerrone MC, Ma JJ, Stephens RS. Cloning and sequence of the gene for the heat shock protein 60 from Chlamydia trachomatis and immunological reactivity of the protein. Infect Immun 1991;59:79-90. 11. Kaufmann SHE. Heat shock proteins and the immune response. Immunol Today 1990; 11:129-36. 12. Res PCM, Thole JER, deVries RRP. Heat shock proteins in immunopathology. Cu~w Opin Immunol 1991;3:924-9. 13. Arno JN, Yuan Y, Cleary RE, Mom'ison RP. Serologic responses of infertile women to the 60-kd chlamydia] heat shock protein (hsp60). Fertil Steril 1995;64:730-5. 14. Witkin SS, Jeremias J, Toth M, Ledger WJ. Cell mediated immune response to the recombinant 57 KD heat shock protein of Chlamydia trachomatis in women with salpingitis. J Infect Dis 1993; 167:1379-83. 15. Witkin SS, Jeremias J, Toth M, Ledger WJ. Proliferative response to conserved epitopes of the Chlamydia trachomatis and human 60-kilodalton heat shock proteins by lymphocytes from women with salpingitis. Am J Obstet Gynecol 1994; 171:455-60. 16. Yi Y, Zhong G, Brunham RC. Continuous B-cell epitopes in Chlamydia trachomatis heat shock protein 60. Infect Immun 1993;61:1117-20. 17. Wang SP, Grayston JT. Human serology in Chlamydia trachomatis infection with microimmunofluorescence test. J Infect Dis 1974; 130:388-97. 18. Newhall WJ, Batteiger B, Jones RB. Analysis of the human serological response to proteins of Chlamydia trachomatis. Infect Immun 1982;38:1181-9. 19. Toye B, Zhong G, Peeling R, Brunham RC. Immunologic characterization of a cloned fi'agment containing the species-specific epitope from the major outer membrane protein of Chlamydia trachomatis. Infect Immun 1990;58:3909-13. 20. Collins JA. Male infertility: the interpretation of the diagnostic assessment. In: Mishell DR, Paulsen CA, Lobo RA, editors. Yearbook olin fertility 1989. Chicago: Year Book Medical Publishers, 1989:45-75. 21. Simel DL, Samsa GP, Matchar DB. Likelihood ratios with confidence: sample size estimation for diagnostic test studies. J Clin Epidemiol 1991;44:763-70. 22. Moore DE, Foy HM, Daling JR, Grayston JT, Spadoni LR, Wang S, et al. Increased frequency of serum antibodies to Chlamydia trachomatis in infertility due to tubal disease. Lancet 1982; 1:574-7. 23. Grayston JT. Infections caused by Chlamydia pneumoniae strain TWAR. J Infect Dis 1992; 15:757-63. 24. Freidank HM, Clad A, Herr AS, Wiedmannalahmad M, Jung B. Immune response to Chlamydia trachomatis heat shock protein in infertile female patients and influence of Chlamydiu pneumoniae antibodies. Eur J Clin Microbiol Infect Dis 1995; 14:1063-9. 25. Claman P, Amimi MN, Peeling RW, Toye B, Jessamine P. Does serologic evidence of remote Chlamydia trachomatis infection and its heat shock protein (CHSP60) affect in vitro fertilization-embryo transfer outcome? Fertil Steril 1996; 65:146-9.
Fertility and Sterility ®
Vol. 67, No. 3, March 1997
Copyright ' 1997 American Snciety for Reproductive Medicine
Printed on acid-free paper in U. S. A.
The presence of serum antibody to the chlamydial heat shock protein (CHSP60) as a diagnostic test for tubal factor infertility*
Paul Claman, M.D.t$ Lisa Honey, M.D.t Rosanna W. Peeling, Ph.D.§
Peter Jessamine, M.D.[[ Baldwin Toye, M.D.¶
Ottawa Civic and General Hospitals, University of Ottawa, Ottawa, Ontario, and the Laboratory Centre for Disease Control, Winnipeg Manitoba, Canada
Objective: To study the utility of testing for heat shock protein 60 (CHSP60) antibodies in the diagnosis of tubal factor infertility. Design: Prospective case control. Setting: Canadian university hospital infertility clinic. P a t i e n t ( s ) : Women presenting for infertility investigation. I n t e r v e n t i o n ( s ) : Sera were collected from 77 patients. Main O u t c o m e Measure(s): The relationship between tubal factor infertility and the presence of antibodies to Chlamydia trachomatis and CHSP60 was assessed. Result(s): There were no significant differences between antibodies to C. trachomatis in women with tubal factor infertility (63%) and other causes of infertility (46%). However, more women with tubal factor infertility (44%) had anti-CHSP60 antibodies compared with other causes of infertility (8%). Antibody testing for C. trachomatis has only a 63% sensitivity and a 54% specificity for detecting tubal factor infertility. In contrast, the CHSP60 antibody test has a 44% sensitivity and a 92% specificity for detecting tubal factor infertility. There is a good positive likelihood ratio of 5.5 for CHSP60 antibody testing detecting the presence of tubal factor infertility. Combining CHSP60 antibody with antibody testing for C. trachomatis has an excellent positive likelihood ratio of 10 for the detection of C. trachomatis-associated tubal factor infertility. Conclusion(s): CHSP60 antibody testing is a more accurate test than antibody testing for C. trachomatis for predicting chlamydia-associated tubal factor infertility. These tests, when used in combination at initial infertility evaluation, would provide a rapid noninterventive means of diagnosing tubal factor infertility. Fertil Steril e 1997;67:501-4 Key Words: Chlamydia heat shock protein 60 (CHSP60), Chlamydia trachomatis, tubal infertility, infertility
Infection with C h l a m y d i a t r a c h o m a t i s is an imp o r t a n t sexually t r a n s m i t t e d disease t h a t h a s been associated with t u b a l infertility in women. The risk
Received July 25, 1996; revised and accepted October 28, 1996. * Supported in part with a grant from Physician Services Incorporated Foundation, and National Laboratory for Sexually Transmitted Diseases, Laboratory Centre for Disease Control, Ottawa, Ontario, Canada. t Division of Reproductive Medicine, Department of Obstetrics and Gynecology,Ottawa Civic Hospital, University of Ottawa. $ Reprint requests: Paul Claman, M.D., 552-737 Parkdale Avenue, Ottawa, Ontario, Canada K1Y4E9 (FAX: 613-724-4942; email: [email protected]). § Laboratory Centre for Disease Control. Vol. 67, No. 3, March 1997
of t u b a l infertility a p p e a r s to i n c r e a s e with the n u m ber a n d severity of episodes of salpingitis r e s u l t i n g in t u b a l occlusion (1). M a n y studies h a v e s h o w n t h a t serologic evidence of prior C h l a r n y d i a infection is s t r o n g l y associated with t u b a l infertility ( 1 - 4 ) . The m e c h a n i s m by w h i c h c h l a m y d i a l infection results in t u b a l d a m a g e h a s been s t u d i e d widely b u t r e m a i n s unclear. P r i m a t e studies s u g g e s t t h a t rein[[Division of Microbiologyand Immunology, Department of Laboratory Medicine, Division of Infectious Disease, Department of Medicine, Ottawa Civic Hospital, University of Ottawa. ¶ Division of Microbiology, Department of Pathology and Laboratory Medicine, Division of Infectious Disease, Department of Medicine, Ottawa General Hospital, University of Ottawa. Claman et al. CHSP60 as a test for tubal infertility
501
fection with C. t r a c h o m a t i s may be the necessary component for chronic salpingitis leading to distal tubal obstruction (5). The immunopathology seen in genital tract infections is similar to those elicited by the chlamydial conjunctival infection leading to scarring trachoma. Reinfection has been identified as an important risk factor in the pathogenesis of trachoma (6). Recent studies have shown a strong association between antibody response to the chlamydia heat shock protein 60 (CHSP60) and ectopic pregnancy (7) as well as the development of chla] mydia-associated tubal infertility (8). Chlamydial heat shock protein is a homolog of the gro-EL family of heat shock proteins (9, 10). This family of proteins is highly conserved among both prokaryotes and eukaryotes (11, 12). It is been suggested that antibodies against conserved epitopes on CHSP60 may cross react with those of h u m a n HSP60 and initiate an autoimmune inflammatory response (7, 13-16). A recent study identified testing for serum antibodies to the whole C. t r a c h o m a t i s organism as more accurate than hysterosalpingography (HSG) in predicting tubal factor infertility (4). We compared the accuracy of chlamydia serology and antibodies against CHSP60 in predicting the diagnosis of tubal factor infertility in a prospective study. MATERIALS AND METHODS
After the proposed study and consent forms were approved by the Research Ethics Committee at the Ottawa Civic Hospital, sera were obtained from 77 unselected women presenting for initial infertility evaluation in the infertility clinic at the University of Ottawa, Ottawa Civic Hospital. All women had standard investigations performed that included BBT charting, and/or midluteal P, and/or late luteal endometrial biopsy, cervical C. t r a c h o m a t i s screening with ELISA or polymerase chain reaction assays. The diagnosis of tubal infertility was made by HSG and/or laparoscopic examination demonstrating distal tubal obstruction or laparoscopic evidence of severe peritubal adhesions. If complete bilateral distal tubal obstruction was found on HSG, a laparoscopic examination was not always performed. All sera were assayed for immunoglobulin (Ig) G and IgM antibody to C. t r a c h o m a t i s by the microimmunofluorescence (MIF) technique of Wang and Grayston (17) using purified Formalin-fixed elementary bodies. Sera were screened at a dilution of 1:8 and titered at twofold dilutions to end point. C h l a m y d i a t r a c h o m a t i s seropositivity was defined as an MIF titer of - 1 : 8 (18). An ELISA using recombinant CHSP60 expressed as a fusion protein with glutathione-S-transferase 502
Claman et al.
CHSP60 as a test for tubal infertility
as an antigen was used to test for the presence of antibody to CHSP60 as described previously (8, 19). Patient sera were diluted 1:500 and incubated with recombinant antigen bound to 96-well microtiter plates. Horseradish peroxidase-conjugated goat anti-human IgG was added and the optical density o f each well was determined. All sera positive by ELISA were confirmed by immunoblotting using recombinant CHSP60 as antigen. Sera were processed blindly and without knowledge of the clinical diagnosis of the patient. Groups were compared with X2 or Fisher's exact test. Odds ratios (OR) with 95% confidence intervals (CI) also were calculated. Likelihood ratios also were computed to help evaluate the diagnostic accuracy of both serologic tests for their ability to predict the presence of tubal disease as the cause of infertility in the patient presenting for infertility evaluation. The calculation of likelihood ratios enables comparison of the diagnostic values of tests independent of prevalence of infection in differing populations. The likelihood ratio of a positive test is calculated as [sensitivity/(100 - specificity)]. A positive likelihood ratio of 2 to 5 indicates a fair clinical test, 5 to 10 indicates a good clinical test, and > 10 is an excellent clinical test. The likelihood ratio of a negative test is calculated as [(100 - sensitivity)/specificity]. A negative likelihood ratio of 0.5 to 0.2 indicates a fair clinical test, 0.2 to 0.1 is a good clinical test, and <0.1 is an excellent clinical test (20). Because likelihood ratios refer to the actual test results before the disease status is known, they are more useful to clinicians than are sensitivity and specificity (21). RESULTS
Seventy-seven women were studied, 16 with the final diagnosis of tubal disease. Two of those with tubal disease also had ovulatory dysfunction, three also had male factor problems and, three had both male factor and ovulatory dysfunction in addition to tubal disease. Sixty-one patients had nontubal causes of infertility: 13 with male factor, 14 with ovulatory dysfunction, with 17 mixed male factor and ovulatory dysfunction, and 17 with idiopathic infertility. There was no difference between the mean age of the women with tubal infertility (32.6 _ 3.2 years) and those with other causes of infertility (31 _+ 4.9 years). A total of six women had a history of pelvic inflammatory disease (PID), but only three of these were found to have tubal disease. Thus, 13 of 16 women with tubal disease in our series had no prior history of PID. Thirteen patients were found to have endometriosis at the time of laparoscopy, and nine of these women had no distortion of tuboovarian anatomy. Seven patients had a prior history Fertility and Sterility ®
of ectopic pregnancy, and three of these women appeared to have normal tubes at laparoscopy after resolution of the ectopic pregnancy (treated with methotrexate or linear salpingostomy). The rest of our patients had a final diagnosis of either male or ovulatory factor infertility. A positive MIF titer was found in 10 of 16 (63%) women with tubal disease and in 28 of 61 (46%) with other causes of infertility (P = 0.2, OR = 1.96, 95% CI = 0.56 to 7.1). Thus, as a screening test for tubal disease, MIF has a sensitivity of 63% and a specificity of 54%, with a positive predictive value of 26% and a negative predictive value of 85%. Calculation of likelihood ratios also show t h a t MIF testing for C. trachomatis antibodies is not a useful clinical test for the prediction of tubal disease in the patient presenting with infertility (positive likelihood ratio = 1.36 and negative likelihood ratio = 0.69). The CHSP60 antibodies were found in 7 of 16 (44%) women with tubal disease and in only 5 of 61 (8%) women with other causes of infertility (P = 0.002, OR = 8.7, 95% CI = 1.9 to 41.9, power = 0.976 for c~ = 0.05). Therefore, as a m a r k e r of tubal infertility, CHSP60 antibody testing has a sensitivity of 44% and a specificity of 92%. For the detection of C. trachomatis-associated tubal disease (defined as tubal disease with a positive MIF titer), the CHSP60 test was positive in 7 of 10 patients with tubal disease and in only 5 of 67 patients with other causes of infertility or with a negative MIF titer (P = 0.0000325, OR = 29, 95%CI = 5 to 215). Thus, the sensitivity of the CHSP60 assay for detecting C. trachomatis associated tubal disease is 70% and its specificity is 93%. The positive predictive value for the CHSP60 test for the diagnosis of tubal infertility is 58% and the negative predictive value is 86%. A patient with tubal disease is 5.5 times more likely than a patient without tubal disease to have antibodies to CHSP60 (positive likelihood ratio). The likelihood ratio of a negative CHSP60 test was 0.61. When both the MIF and CHSP60 test are used, the CHSP60 has a positive likelihood ratio of 10 and a negative likelihood ratio of 0.32, making it an excellent test for the detection of C. trachomatis-associated tubal disease. DISCUSSION
It has been suggested t h a t C. trachomatis antibody testing might be useful as a screening tool in predicting tubal factor infertility (22). As a predictor of tubal disease, we found t h a t the MIF for C. trachomatis test has a very limited specificity of only 54% and a sensitivity of 63% for predicting tubal disease in our clinic population. Positive and negative likelihood ratios of 1.36 and 0.69 show t h a t the MIF is a Vol. 67, No. 3, March 1997
poor clinical test for the detection of tubal disease. Dabekausen et al. (4) have suggested t h a t C. trachom a t i s antibody may be better at diagnosing tubal disease than hysterosalpingography. The antibody assay system used in their study m a y have limited specificity to differentiate between antibodies to C. trachomatis and C. p n e u m o n i a e . The seroprevalence of C. p n e u m o n i a e in most adult populations is 40% to 60% (23). The MIF test allows more specific detection of antibodies against different chlamydial species (24). Furthermore, Dabekausen et al.'s (4) definition of tubal disease included tuboperitoneal adhesions and even unilateral tubal pathology seen at laparoscopy, whereas in our series we defined tubal disease as the presence of bilateral distal tubal obstruction or severe peritubal adhesions. In this prospective study, antibody to CHSP60 was found in 44% of women with tubal factor infertility and only 8% of women with other causes of infertility. Antibody to CHSP60 was found to be a very specific predictor of the diagnosis of chlamydia-associated tubal factor infertility. Although the sensitivity of this test for detecting tubal disease is poor, the 92% specificity of this test for detecting tubal disease makes it an important assay. The positive likelihood ratio of CHSP60 antibody testing for detecting the presence of tubal disease in an unselected population of patients presenting for infertility was 5.5, making it a good clinical test for predicting the presence of tubal factor infertility. These data in this prospective study are consistent with our findings from a retrospective study of CHSP60 antibody response in patients undergoing IVF-ET t r e a t m e n t (8). The sensitivity of any test for the diagnosis of C. trachomatis-associated tubal infertility is dependent on the prevalence of prior genital chlamydial infection in the population being tested. Thus, the CHSP60 antibody test may be limited by the fact that there were patients with other causes of tubal infertility, including previous gonococcus infection, severe endometriosis, tuberculosis, and history of ectopic pregnancies. The CHSP60 antibody testing will detect only tubal damage resulting from C. trachomatis infections, which will not account for tubal disease due to other causes. The sensitivity of the CHSP60 test for detecting tubal disease in C. trachomatis MIF seropositive patients was 70% and the specificity was 93% in this study. If both the MIF and CHSP60 tests are used, the CHSP60 will have a positive likelihood ratio of 10 and a negative likelihood ratio of 0.32, making it an excellent test for .the detection of C. trachomatis-associated tubal infertility (defined as tubal infertility with a positive MIF titer). B r u n h a m et al. (7) previously reported t h a t 19 of 21 (91%) patients with ectopic pregnancy and seropositive for C. trachomatis had serum antibodies to Claman et al. CHSP60 as a test for tubal infertility
503
CHSP60. Seven women in our population sample had a prior history of ectopic pregnancy, four of whom were MIF positive. Three of these four women had CHSP60 antibodies (75%). Only four of these seven women had obvious tubal disease. One of the important reasons for testing for chlamydia as opposed to other sexually transmitted diseases as a cause of tubal disease and infertility is that the majority of genital chlamydia infections in women are asymptomatic. Undetected and untreated these infections can ascend into the upper genital tract and cause PID with resultant ectopic pregnancy arid tubal occlusion. Many studies have shown that most women with tubal infertility did not recall a history of chlamydia infection (3). These tests would be a rapid means of identifying these women and of establishing the diagnosis without other unnecessary and costly interventive procedm'es. The usefulness of the CHSP60 antibody test lies in its high specificity (92%) and good positive likelihood ratio (5.5) for the presence of tubal factor infertility. A negative MIF test has a very good (85%) negative predictive value, which makes it useful alone in ruling out tubal factor as the cause of infertility. A positive MIF assay combined with CHSP60 antibody testing makes for an excellent test ofC. trachomatisassociated tubal infertility (positive likelihood ratio = 10). The CHSP60 antibody test combined with MIF testing should be introduced as a diagnostic tool as part of the initial infertility evaluation. Patients with a positive CHSP60 antibody assay may even consider moving into IVF-ET treatment rather than pursue further interventive and expensive diagnostic testing, especially because IVF-ET appears to be particularly effective in the treatment of tubal factor infertility patients who have antibodies to the CHSP60 (25). REFERENCES 1. Westrom L. Chlamydia and its effect on reproduction. J Br Fertil Soc 1998; 1:23-30. 2. Brunham RC, MacLean IW, Binns B, Peeling RW. Chlumydia trachomatis: its role in tubal infertility. J Infect Dis 1985; 152:1275-82. 3. Cates W Jr, Wasserheit JN. Genital chlamydial infections: epidemiology and reproductive sequelae. Am J Obstet Gynecol 1991; 164:1771-81. 4. Dabekausen YAJM, Evers JLH, Land JA, Stals FS. Chlamydia trachomatis antibody testing is more accurate than hysterosalloingography in predicting tubal factor infertility. Fertil Steril 1994;61:833-7. 5. Patton DL, Kuo CC, Wang SP, Halbert SA. Distal tubal obstruction induced by repeated Chlamydia trachomatis salpingeal infections in pig-tailed macaques. J Infect Dis 1987; 155:1292-9. 6. Grayson JT, Wang SP, Yeh hi, Kuo CC. Importance of reinfection in the pathogenesis of trachoma. Rev Infect Dis 1985;7:717-25.
504
Claman et al.
CHSP60 as a test for tubal infertility
7. Brunham RC, Peeling R, MacLean I, Kosseim ML, Paraskevas M. Chlamydia trachomatis-associated ectopic pregnancy: serologic and histologic correlates. J Infect Dis 1992; 165:1076-81. 8. Toye B, Laferriere C, Claman P, Jessamine P, Peeling R. Association between antibody to the chlamydial heat-shock protein and tubal infertility. J Infect Dis 1993; 168:1236-40. 9. Morrison RP, Su H, Lyng K, Yran Y. The Chlamydia trachomatis Hyp operon is homologous to the groE stress response operon of Esherichia coli. Infect Immun 1990;58:2701-5. 10. Cerrone MC, Ma JJ, Stephens RS. Cloning and sequence of the gene for the heat shock protein 60 from Chlamydia trachomatis and immunological reactivity of the protein. Infect Immun 1991;59:79-90. 11. Kaufmann SHE. Heat shock proteins and the immune response. Immunol Today 1990; 11:129-36. 12. Res PCM, Thole JER, deVries RRP. Heat shock proteins in immunopathology. Cu~w Opin Immunol 1991;3:924-9. 13. Arno JN, Yuan Y, Cleary RE, Mom'ison RP. Serologic responses of infertile women to the 60-kd chlamydia] heat shock protein (hsp60). Fertil Steril 1995;64:730-5. 14. Witkin SS, Jeremias J, Toth M, Ledger WJ. Cell mediated immune response to the recombinant 57 KD heat shock protein of Chlamydia trachomatis in women with salpingitis. J Infect Dis 1993; 167:1379-83. 15. Witkin SS, Jeremias J, Toth M, Ledger WJ. Proliferative response to conserved epitopes of the Chlamydia trachomatis and human 60-kilodalton heat shock proteins by lymphocytes from women with salpingitis. Am J Obstet Gynecol 1994; 171:455-60. 16. Yi Y, Zhong G, Brunham RC. Continuous B-cell epitopes in Chlamydia trachomatis heat shock protein 60. Infect Immun 1993;61:1117-20. 17. Wang SP, Grayston JT. Human serology in Chlamydia trachomatis infection with microimmunofluorescence test. J Infect Dis 1974; 130:388-97. 18. Newhall WJ, Batteiger B, Jones RB. Analysis of the human serological response to proteins of Chlamydia trachomatis. Infect Immun 1982;38:1181-9. 19. Toye B, Zhong G, Peeling R, Brunham RC. Immunologic characterization of a cloned fi'agment containing the species-specific epitope from the major outer membrane protein of Chlamydia trachomatis. Infect Immun 1990;58:3909-13. 20. Collins JA. Male infertility: the interpretation of the diagnostic assessment. In: Mishell DR, Paulsen CA, Lobo RA, editors. Yearbook olin fertility 1989. Chicago: Year Book Medical Publishers, 1989:45-75. 21. Simel DL, Samsa GP, Matchar DB. Likelihood ratios with confidence: sample size estimation for diagnostic test studies. J Clin Epidemiol 1991;44:763-70. 22. Moore DE, Foy HM, Daling JR, Grayston JT, Spadoni LR, Wang S, et al. Increased frequency of serum antibodies to Chlamydia trachomatis in infertility due to tubal disease. Lancet 1982; 1:574-7. 23. Grayston JT. Infections caused by Chlamydia pneumoniae strain TWAR. J Infect Dis 1992; 15:757-63. 24. Freidank HM, Clad A, Herr AS, Wiedmannalahmad M, Jung B. Immune response to Chlamydia trachomatis heat shock protein in infertile female patients and influence of Chlamydiu pneumoniae antibodies. Eur J Clin Microbiol Infect Dis 1995; 14:1063-9. 25. Claman P, Amimi MN, Peeling RW, Toye B, Jessamine P. Does serologic evidence of remote Chlamydia trachomatis infection and its heat shock protein (CHSP60) affect in vitro fertilization-embryo transfer outcome? Fertil Steril 1996; 65:146-9.
Fertility and Sterility ®