- Email: [email protected]
Chlamydia trachomatis antibody testing is more accurate than hysterosalpingography in predicting tubal factor infertility*
Vol. 61, No.5, May 1994
FERTILITY AND STERILITY
Printed on acid-free paper in U. S. A.
Copyright c 1994 The American Fertility Society
Chlamydia trachomatis antibody testing is more accurate than hysterosalpingography in predicting tubal factor infertility·
Yvonne A. J. M. Dabekausen, M.D.t Johannes L. H. Evers, M.D.t Jolande A. Land, M.D.t:J: Frans S. Stals, M.D.§ Academisch Ziekenhuis Maastricht and the University of Limburg, Maastricht, The Netherlands
Objective: To compare the likelihood of abnormal Chlamydia trachomatis antibody test results with that of abnormal hysterosalpingography (HSG) test results in patients with tubal factor infertility. Design: Anti-C. trachomatis immunoglobulin G antibodies were determined prospectively in 211 consecutive infertility patients by means of an indirect fluorescent antibody technique. The results were compared with the results of HSG with respect to their predictive value of tubal factor infertility. Likelihood ratio calculations were used. Setting: University hospital-based, tertiary care infertility clinic. Patients: A series of 211 consecutive infertility patients. Interventions: C. trachomatis antibody testing, HSG, laparoscopy. Main Outcome Measures: Likelihood ratios for abnormal C. trachomatis antibody test results and abnormal HSG results in infertility patients, as assessed by laparoscopy. Results: The positive likelihood ratio for C. trachomatis antibody testing was 9.1, indicating a patient with tubal factor infertility to be 9.1 times more likely to have abnormal serology results than a patient without tubal factor infertility. This was superior to HSG, which had a positive likelihood ratio of 2.6 in our study and of 1.6 to 6.1 in the literature. The odds ratio of C. trachomatis antibody testing was 31.5 in our study. Its 90% confidence interval (8.3 to 138.5) did not overlap that of HSG as calculated from a meta-analysis of literature reports (5.3 to 7.9). Conclusions: C. trachomatis antibody testing is simple, inexpensive, and causes minimal inconvenience to the patient. It is more likely than HSG to be abnormal in patients with tubal factor infertility. C. trachomatis antibody testing deserves to become an integral component of the initial Fertil Steril1994;61:833-7 fertility work-up. Key Words: Chlamydia trachomatis antibody, hysterosalpingography, tubal factor infertility, likelihood ratio, odds ratio
Infertility affects approximately 10% of couples. Fourteen percent (1) to 38% (2) of female infertility
Received September 13, 1993; revised and accepted January 12,1994. * Presented in part at the 49th Annual Meeting of The American Fertility Society, Montreal, Quebec, Canada, October 9 to 14,1993. t Department of Obstetrics and Gynaecology. :j: Reprint requests: Jolande A. Land, M.D., Department of Obstetrics and Gynaecology, Academisch Ziekenhuis Maastricht, P.O. Box 5800, 6202 AZ Maastricht, The Netherlands (FAX: +31-43-874765). § Department of Medical Microbiology. Vol. 61, No.5, May 1994
is associated with a "tubal factor." The single most important cause of tubal pathology is inflammatory disease. The incidence of genital infections with Chlamydia trachomatis is increasing worldwide (3). In 50% to 80% of the female patients the course of this infection is asymptomatic (4). C. trachoma tis causes damage to the ciliated epithelium of the tubes, compromising tubal transport and increasing the risk for ectopic pregnancy and infertility. Current C. trachomatis infections can be detected by culturing or staining a direct specimen. Cultures, however, can be false-negative, especially in ascending infections of the genital tract (5).
Dabekausen et al.
Chlamydia antibody and tubal infertility
833
Extinct C. trachomatis infections can be revealed still by serologic tests (5) because C. trachomatis antibodies persist after pelvic inflammatory disease (6). Because the majority of C. trachomatis infections remain unnoticed by the patient but give rise to tuboperitoneal pathology and persistent antibody formation, serologic C. trachomatis antibody testing has been introduced into the infertility work-up (7). The first aim of this study was to investigate the prevalence of C. trachomatis antibodies in serum in an infertility population and to determine their suitability in screening for tuboperitoneal disease, as found at laparoscopy. The second aim was to compare the diagnostic test properties of serum C. trachomatis antibody testing with those of hysterosalpingography (HSG) with respect to their predicting tubal factor infertility. C. trachomatis antibody testing and HSG were compared by calculating their respective odds and likelihood ratios (8).
MATERIALS AND METHODS
Two hundred eleven female infertility patients who presented consecutively to our infertility clinic participated in a prospective study. Blood was drawn at their initial visit and an indirect microimmunofluorescent antibody technique for C. trachomatis immunoglobulin (Ig) G antibodies (9, 10) was used. In short, 20 JLL of serum was diluted eight times in phosphate-buffered saline (PBS), pH 7.4, and incubated on the C. trachomatis-spot immunofluorescence substitute slides (Egg-grown C. trachomatis biovar L2; BioMerieux, 's Hertogenbosch, The Netherlands) for 30 minutes at 37°C in a moist chamber. The slides were washed three times for 5 minutes in PBS and incubated with fluoresceinconjugated rabbit anti-human IgG (Dako; ITK Diagnostics B.V., Uithoorn, The Netherlands) diluted in PBS for 30 minutes at 37°C. After three washings in PBS and one in ultrapure water processed through a milli-Q purifying system (Millipore, Bedford, MA), slides were embedded in Fluoprep mounting medium (BioMerieux) under a coverslip. A positive reaction is characterized by specific fluorescence of the C. trachomatis elementary body. For a quantitative determination, serial dilution in PBS was performed. After a follow-up period of 6 to 18 months, both an HSG and a laparoscopy had been performed in
834
Dabekausen et aI.
34 patients; 37 additional patients had undergone only a laparoscopy. Therefore, results were available to compare C. trachomatis antibody tests with laparoscopy findings for 71 patients, and HSG results could be compared with the findings at laparoscopy for 34 patients. All HSGs were performed in the proliferative phase of the menstrual cycle with oil-soluble contrast medium (Lipiodol Ultrafluide; Laboratoire Guerbet, Aulnay-sous-bois, France). An HSG was considered abnormal if one or both tubes did not allow passage of contrast medium. Peritubal adhesions were suspected whenever the contrast medium failed to spread freely in the peritoneal cavity and/or pockets of contrast medium were apparent on the delayed picture. Laparoscopy was performed in the proliferative phase of the cycle. Tubal testing was done with methylene blue dye. Patients were classified as having tuboperitoneal abnormalities if evidence of fimbrial, peritubal, and/or periovarian adhesions; obstruction of one or both fallopian tubes; hydrosalpinx; or a combination of these was encountered during laparoscopy. At the time oflaparoscopy and HSG, the investigators were blind to the Chlamydia serology results. The most suitable clinical cutoff level of the antiC. trachoma tis antibody titer in serum was calculated by means of construction of a receiver operating characteristics curve (11). An antibody titer of 8 appeared to be the optimal cutoff level. Consequently, patients in whom a titer of >8 was found were considered seropositive. The diagnostic value of the C. trachomatis antibody test was compared with the value of HSG in predicting tubal factor infertility in the individual patient using likelihood ratios. Likelihood ratios, in contrast to the more conventionally used positive and negative predictive values, are not affected by the prevalence of disease in the population studied. Therefore, they can be used to compare the outcome of the same test in different populations and to compare various tests of the same disease entity in the same population. The likelihood ratio of a positive test result (LR+) indicates the likelihood of a positive test in a patient with the disease over the likelihood of a positive test in a patient without the disease. The LR- indicates the likelihood of a negative test in a patient with the disease over the likelihood of a negative test in a patient without the disease. The LR+ is calculated as [sensitivity/(1 - specificity)]. The LR- is calculated as [(1 - sen-
Chlamydia antibody and tubal infertility
Fertility and Sterility
Table 1 Tuboperitoneal Abnormalities at Laparoscopy in Seropositive and Seronegative* Infertility Patients
C. trachomatis antibody testing
Abnormal laparoscopy
Normal laparoscopy
Positive Negative
25t
3 34
9
* C. trachomatis antibody titer> 8 for seropositive patients; :$8 for seronegative patients. t p < 0.000001 (X 2 test); sensitivity: 74%; specificity: 92%; LR+: 9.1; X2 LR-: 0.3; OR: 31.5 (90% confidence interval: 8.3 to 138.5).
sitivity)/specificity). Calculation of LRs yields a score that allows categorization of test results: an LR+ between 2 and 5 indicates a fair clinical test, between 5 and 10 is good, and> 10 is excellent. An LR- between 0.5 and 0.2 indicates a fair clinical test, between 0.2 and 0.1 is good, and <0.1 is excellent (8). The odds ratio (OR) can be calculated from the LRs: OR = LR+ /LR- (12). The OR reflects the probability of a patient with an abnormal test to have tubal factor infertility. The OR equals 1 if the level of the test equals the prevalence of the disease. When the level exceeds the prevalence the OR is >1 (12). We calculated the LR+, LR-, and OR for C. trachomatis antibody testing and for HSG, both for our own study and for those literature reports allowing their calculation. The LRs and the OR for C. trachomatis antibody testing in our study were compared with the LRs and the OR for HSG, both from our study and from the literature.
RESULTS
In 60 of 211 patients studied (28.5%), the C. trachomatis antibody test was positive (anti-C. trachomatis IgG titer> 8). In 71 of 211 patients, a laparoscopy was performed (Table 1): 25 of the 28 seropositive patients (89%) had tuboperitoneal abnormalities. Of the 43 seronegative patients, only 9 (21 %) had adhesions or tubal obstruction. This difference is statistically highly significant (P < 0.000001; X 2 test). In 34 of 211 patients, an HSG as well as a laparoscopy was performed. In 24 of these 34 patients the results corresponded, but in 10 patients a discrepancy between HSG and laparoscopy was found (P = 0.06; Fisher's exact test) (Table 2). According to Bayes' theorem, the probability of tubal factor infertility for a patient in our popula-
Vol. 61, No.5, May 1994
tion is 88% if she has abnormal C. trachomatis antibody testing and 59% if she has an abnormal HSG. To obtain a prevalence independent measure of the reliability of C. trachomatis antibody testing, likelihood ratios were calculated. The LR+ of C. trachomatis antibody testing was 9.1, indicating a patient with tubal factor infertility to be 9.1 times more likely to have a positive test result (i.e., titer > 8) than a patient without tubal factor infertility. In comparison, the LR+ for HSG in the same group of patients was only 2.6. The LR - for C. trachomatis antibody testing was 0.3, indicating a patient with tubal factor infertility to be 0.3 times as likely to have a negative test (i.e., titer::;; 8) as a patient without the disease. The LR- for HSG was 0.5. The respective ORs were 31.5 for C. trachomatis antibody testing and 4.8 for HSG.
DISCUSSION
In the infertility patient, laparoscopy is the most accurate method to diagnose tuboperitoneal pathology but, for screening purposes, HSG is used widely. Hysterosalpingography is a disagreeable, annoying test to the patient. It may be painful and has a high false-positive and false-negative rate. The false-positive outcome (in most studies around 25 %) (13-23) is thought to result from tubal spasm, dissimilar tubal filling pressure, too high viscosity of the contrast medium, and faulty technique. False-negative findings have been reported in around 40% (13-23) and are considered to be due to peritubal adhesions not visualized on the delayed picture. We investigated the value of serum C. trachomatis antibody testing in screening for tubal factor infertility. With laparoscopy as a gold standard, we compared the predictive value of the C. trachomatis antibody titer with HSG. For this comparison we used LRs, prevalence independent test characteristics.
Table 2 Tuboperitoneal Abnormalities at Laparoscopy in Infertility Patients with Abnormal and Normal HSG
HSG
Abnormal laparoscopy
Normal laparoscopy
Abnormal Normal
7* 5
5 17
* p = 0.06 (Fisher's exact test); sensitivity: 58%; specificity: 77%; LR+: 2.6; LR-: 0.5; OR: 4.8 (90% confidence interval: 1.0 to 21.8).
Dabekausen et al.
Chlamydia antibody and tubal infertility
835
Table 3 LR+ and LR- of HSGs and Corresponding ORs in Predicting Tuboperitoneal Abnormalities as Calculated From Literature Reports Studies
LR+
LR-
OR*
Snowden et aI., 1984 (13) Moghissi and Kim, 1975 (14) Keirse and Vandervellen, 1973 (15) Gabos, 1976 (16) Swolin and Rosencrantz, 1972 (17) Present study Cumming and Taylor, 1980 (18) Ismajovich et aI., 1986 (19) Philipsen and Hansen, 1981 (20) World Health Organization, 1986 (21) Maathuis et aI., 1972 (22) Reshef et aI., 1989 (23)
6.1 4.9 4.6 2.9 2.7 2.6 2.3 2.0 2.0 1.8 1. 7 1.6
0.3 0.3 0.3 0.4 0.2 0.5 0.4 0.6 0.3 0.3 0.5 0.3
18.9 20.4 16.6 4.3 12.5 4.8 6.3 3.4 7.3 5.6 3.3 6.1
* Cumulative OR (90% confidence interval): 6.4 (5.3 to 7.9).
The LR+ of C. trachomatis antibody testing was 9.1; the LR+ ofthe HSG in this study was only 2.6. Theoretically, a poor performance of HSG in our clinic might be a reason for this disappointing test outcome. A review of the literature, however, showed LR+ of the HSG to range from 1.6 to 6.1 in those reports giving sufficient details for their calculation (Table 3). Even when the best LR+ of HSG in literature is compared with the LR+ ofthe C. trachomatis antibody test in our study, the latter reflects tuboperitoneal abnormalities better than HSG. The LR- of C. trachomatis antibody testing is 0.3 in our study and is comparable to the HSG studies with the best discriminatory value reported in the literature (Table 3). The group of patients who had HSG results available in our study was too small to allow for a reliable positioning of the new test, C. trachomatis antibody testing, in the diagnostic armamentarium to which the old test, HSG, belongs. Therefore, we estimated the cumulative OR and its confidence limits for HSG from a meta-analysis of those literature reports that provided sufficient details for their calculation. The cumulative OR for HSG was 6.4 with 90% confidence limits of 5.3 and 7.9. The OR of C. trachoma tis antibody testing differed significantly with an 0 R of 31.5 and 90% confidence limits of 8.3 and 138.5. Of the seronegative patients in our study, 9 of 43 (21 %) had tuboperitoneal abnormalities. These were due to previous pelvic surgery, endometriosis, or pelvic inflammatory disease caused by microorganisms other than C. trachomatis. In the present
836
Dabekausen et al.
study, the likelihood of disease increased with the height of the titer. All patients with a titer> 128 had tuboperitoneal disease. These results are in accordance with the findings by Minassian and Wu (24), who reported a correlation between the height of the titer and the severity of pelvic adhesions. Toye and co-workers (25) recently published a study of the association between antibody to the chlamydial heat-shock protein and tubal infertility. Their results suggest that chlamydial heat-shock protein determination allows further differentiation within the group of C. trachomatis-positive patients between those with and those without tubal damage. In conclusion, our study focused on serum anti -C. trachomatis IgG antibody screening in women presenting with infertility. C. trachomatis antibody testing by microimmunofluorescence is simple, inexpensive (Dfl 24 [$13] compared with Dfl 323 [$179] in The Netherlands) and causes minimal inconvenience to the patient. It is more likely than HSG to give abnormal results in patients with tubal factor infertility. C. trachomatis antibody testing deserves to become an integral component of the initial fertility work-up.
Chlamydia antibody and tubal infertility
REFERENCES 1. Hull MGR, Glazener CMA, Kelly NJ, Conway DI, Foster P A, Hinton RA, et al. Population study of causes, treatment, and outcome of infertility. Br Med J 1985;291:1693-7. 2. Arronet GH, Eduljee SY, O'Brien JR. A nine·year survey of fallopian tube dysfunction in human infertility. Fertil Steril 1969;20:903-18. 3. Paavonen J, Wolner-Hanssen P. Chlamydia trachomatis: a major threat to reproduction. Hum Reprod 1989;4:111-24. 4. Thejls H, Rahm VA, Rosen G, Gnarpe H. Correlation between chlamydia infection and clinical evaluation, vaginal wet smear, and cervical swab test in female adolescents. Am J Obstet Gynecol 1987;157:974-6. 5. Schoenwald E, Schmidt BL, Steinmetz G, Hosmann J, Pohla·Gubo G, Luger A, et al. Diagnosis of Chlamydia trachomatis infections. Culture versus serology. Eur J Epide· miol 1988;4:75-82. 6. Puolakkainen M, Vesterinen E, Purola E, Saikku P, Paavonen J. Persistence of Chlamydial antibodies after pelvic inflammatory disease. J Clin Microbiol 1986;23:924-8. 7. 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. 8. Collins JA. Male infertility: the interpretation of the diagnostic assessment. In: Mishell DR, Paulsen CA, Lobo RA, editors. Yearbook of infertility 1989. Chicago: Year Book Medical Publishers, 1989:45-75. 9. Treharne JD, Darougar S, Jones BR. Modification of the micro-immunofluorescence test to provide a routine sero-
Fertility and Sterility
diagnostic test for chlamydial infection. J Clin Pathol 1977;30:510-7. 10. Barnes RC. Laboratory diagnosis of human chlamydial infections. Clin Microbiol Rev 1989;2:119-37. 11. McNeil BJ, Hanley JA. Statistical approaches to the analysis of receiver operating characteristics. Med Decis Making 1984;4:137-50. 12. Kraemer HC. Evaluating medical tests, objective and quantitative guidelines. Newbury Park: Sage Publications, 1992:103-13. 13. Snowden EU, Jarrett JC II, Dawood MY. Comparison of diagnostic accuracy of laparoscopy, hysteroscopy, and hysterosalpingography in evaluation of female infertility. Fertil Steril1984;41:709-13. 14. Moghissi KS, Sim GS. Correlation between hysterosalpingography and pelvic endoscopy for the evaluation of tubal factor. Fertil Steril 1975;26:1178-81. 15. Keirse MJ CN, Vandervellen R. A comparison of hysterosalpingography and laparoscopy in the investigation of infertility. Obstet Gynecol 1973;41:685-8. 16. Gabos P. A comparison of hysterosalpingography andendoscopy in evaluation of tubal function in infertile women. Fertil Steril 1976;27:238-42. 17. Swolin K, Rosencrantz M. Laparoscopy vs. hysterosalpingography in sterility investigations: a comparative study. Fertil Steril 1972;23:270-3. 18. Cumming DC, Taylor PJ. Combined laparoscopy and hys-
Vol. 61, No.5, May 1994
19.
20.
21.
22.
23.
24.
25.
teroscopy in the investigation of the ovulatory infertile female. Fertil SteriI1980;33:475-8. Ismajovich B, Wexler S, Golan A, Langer L, David MP. The accuracy of hysterosalpingography versus laparoscopy in evaluation of infertile women. Int J Gynaecol Obstet 1986;24:9-12. Philipsen T, Hansen BB. Comparative study of hysterosalpingography and laparoscopy in infertile patients. Acta Obstet Gynecol Scand 1981;60:149-51. WHO study. Comparative trial of tubal insufflation, hysterosalpingography, and laparoscopy with dye hydrotubation for assessment of tubal patency. Fertil Steril1986;46:11017. Maathuis JB, Horbach JGM, van Hall EV. A comparison of the results of hysterosalpingography and laparoscopy in the diagnosis of fallopian tube dysfunction. Fertil Steril 1972;23:428-31. Reshef E, Daniel WW, Foster JC, Bradley EL, Blackwell RE, Younger JB. Comparison between I-hour and 24-hour follow-up radiographs in hysterosalpingography using oil based contrast media. Fertil Steril 1989;52:753-5. Minassian SS, Wu CH. Chlamydia antibody by enzymelinked immunosorbent assay and associated severity of tubal factor infertility. Fertil Steril1992;58:1245-7. Toye B, Caferriere 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.
Dabekausen et al.
Chlamydia antibody and tubal infertility
837
FERTILITY AND STERILITY
Printed on acid-free paper in U. S. A.
Copyright c 1994 The American Fertility Society
Chlamydia trachomatis antibody testing is more accurate than hysterosalpingography in predicting tubal factor infertility·
Yvonne A. J. M. Dabekausen, M.D.t Johannes L. H. Evers, M.D.t Jolande A. Land, M.D.t:J: Frans S. Stals, M.D.§ Academisch Ziekenhuis Maastricht and the University of Limburg, Maastricht, The Netherlands
Objective: To compare the likelihood of abnormal Chlamydia trachomatis antibody test results with that of abnormal hysterosalpingography (HSG) test results in patients with tubal factor infertility. Design: Anti-C. trachomatis immunoglobulin G antibodies were determined prospectively in 211 consecutive infertility patients by means of an indirect fluorescent antibody technique. The results were compared with the results of HSG with respect to their predictive value of tubal factor infertility. Likelihood ratio calculations were used. Setting: University hospital-based, tertiary care infertility clinic. Patients: A series of 211 consecutive infertility patients. Interventions: C. trachomatis antibody testing, HSG, laparoscopy. Main Outcome Measures: Likelihood ratios for abnormal C. trachomatis antibody test results and abnormal HSG results in infertility patients, as assessed by laparoscopy. Results: The positive likelihood ratio for C. trachomatis antibody testing was 9.1, indicating a patient with tubal factor infertility to be 9.1 times more likely to have abnormal serology results than a patient without tubal factor infertility. This was superior to HSG, which had a positive likelihood ratio of 2.6 in our study and of 1.6 to 6.1 in the literature. The odds ratio of C. trachomatis antibody testing was 31.5 in our study. Its 90% confidence interval (8.3 to 138.5) did not overlap that of HSG as calculated from a meta-analysis of literature reports (5.3 to 7.9). Conclusions: C. trachomatis antibody testing is simple, inexpensive, and causes minimal inconvenience to the patient. It is more likely than HSG to be abnormal in patients with tubal factor infertility. C. trachomatis antibody testing deserves to become an integral component of the initial Fertil Steril1994;61:833-7 fertility work-up. Key Words: Chlamydia trachomatis antibody, hysterosalpingography, tubal factor infertility, likelihood ratio, odds ratio
Infertility affects approximately 10% of couples. Fourteen percent (1) to 38% (2) of female infertility
Received September 13, 1993; revised and accepted January 12,1994. * Presented in part at the 49th Annual Meeting of The American Fertility Society, Montreal, Quebec, Canada, October 9 to 14,1993. t Department of Obstetrics and Gynaecology. :j: Reprint requests: Jolande A. Land, M.D., Department of Obstetrics and Gynaecology, Academisch Ziekenhuis Maastricht, P.O. Box 5800, 6202 AZ Maastricht, The Netherlands (FAX: +31-43-874765). § Department of Medical Microbiology. Vol. 61, No.5, May 1994
is associated with a "tubal factor." The single most important cause of tubal pathology is inflammatory disease. The incidence of genital infections with Chlamydia trachomatis is increasing worldwide (3). In 50% to 80% of the female patients the course of this infection is asymptomatic (4). C. trachoma tis causes damage to the ciliated epithelium of the tubes, compromising tubal transport and increasing the risk for ectopic pregnancy and infertility. Current C. trachomatis infections can be detected by culturing or staining a direct specimen. Cultures, however, can be false-negative, especially in ascending infections of the genital tract (5).
Dabekausen et al.
Chlamydia antibody and tubal infertility
833
Extinct C. trachomatis infections can be revealed still by serologic tests (5) because C. trachomatis antibodies persist after pelvic inflammatory disease (6). Because the majority of C. trachomatis infections remain unnoticed by the patient but give rise to tuboperitoneal pathology and persistent antibody formation, serologic C. trachomatis antibody testing has been introduced into the infertility work-up (7). The first aim of this study was to investigate the prevalence of C. trachomatis antibodies in serum in an infertility population and to determine their suitability in screening for tuboperitoneal disease, as found at laparoscopy. The second aim was to compare the diagnostic test properties of serum C. trachomatis antibody testing with those of hysterosalpingography (HSG) with respect to their predicting tubal factor infertility. C. trachomatis antibody testing and HSG were compared by calculating their respective odds and likelihood ratios (8).
MATERIALS AND METHODS
Two hundred eleven female infertility patients who presented consecutively to our infertility clinic participated in a prospective study. Blood was drawn at their initial visit and an indirect microimmunofluorescent antibody technique for C. trachomatis immunoglobulin (Ig) G antibodies (9, 10) was used. In short, 20 JLL of serum was diluted eight times in phosphate-buffered saline (PBS), pH 7.4, and incubated on the C. trachomatis-spot immunofluorescence substitute slides (Egg-grown C. trachomatis biovar L2; BioMerieux, 's Hertogenbosch, The Netherlands) for 30 minutes at 37°C in a moist chamber. The slides were washed three times for 5 minutes in PBS and incubated with fluoresceinconjugated rabbit anti-human IgG (Dako; ITK Diagnostics B.V., Uithoorn, The Netherlands) diluted in PBS for 30 minutes at 37°C. After three washings in PBS and one in ultrapure water processed through a milli-Q purifying system (Millipore, Bedford, MA), slides were embedded in Fluoprep mounting medium (BioMerieux) under a coverslip. A positive reaction is characterized by specific fluorescence of the C. trachomatis elementary body. For a quantitative determination, serial dilution in PBS was performed. After a follow-up period of 6 to 18 months, both an HSG and a laparoscopy had been performed in
834
Dabekausen et aI.
34 patients; 37 additional patients had undergone only a laparoscopy. Therefore, results were available to compare C. trachomatis antibody tests with laparoscopy findings for 71 patients, and HSG results could be compared with the findings at laparoscopy for 34 patients. All HSGs were performed in the proliferative phase of the menstrual cycle with oil-soluble contrast medium (Lipiodol Ultrafluide; Laboratoire Guerbet, Aulnay-sous-bois, France). An HSG was considered abnormal if one or both tubes did not allow passage of contrast medium. Peritubal adhesions were suspected whenever the contrast medium failed to spread freely in the peritoneal cavity and/or pockets of contrast medium were apparent on the delayed picture. Laparoscopy was performed in the proliferative phase of the cycle. Tubal testing was done with methylene blue dye. Patients were classified as having tuboperitoneal abnormalities if evidence of fimbrial, peritubal, and/or periovarian adhesions; obstruction of one or both fallopian tubes; hydrosalpinx; or a combination of these was encountered during laparoscopy. At the time oflaparoscopy and HSG, the investigators were blind to the Chlamydia serology results. The most suitable clinical cutoff level of the antiC. trachoma tis antibody titer in serum was calculated by means of construction of a receiver operating characteristics curve (11). An antibody titer of 8 appeared to be the optimal cutoff level. Consequently, patients in whom a titer of >8 was found were considered seropositive. The diagnostic value of the C. trachomatis antibody test was compared with the value of HSG in predicting tubal factor infertility in the individual patient using likelihood ratios. Likelihood ratios, in contrast to the more conventionally used positive and negative predictive values, are not affected by the prevalence of disease in the population studied. Therefore, they can be used to compare the outcome of the same test in different populations and to compare various tests of the same disease entity in the same population. The likelihood ratio of a positive test result (LR+) indicates the likelihood of a positive test in a patient with the disease over the likelihood of a positive test in a patient without the disease. The LR- indicates the likelihood of a negative test in a patient with the disease over the likelihood of a negative test in a patient without the disease. The LR+ is calculated as [sensitivity/(1 - specificity)]. The LR- is calculated as [(1 - sen-
Chlamydia antibody and tubal infertility
Fertility and Sterility
Table 1 Tuboperitoneal Abnormalities at Laparoscopy in Seropositive and Seronegative* Infertility Patients
C. trachomatis antibody testing
Abnormal laparoscopy
Normal laparoscopy
Positive Negative
25t
3 34
9
* C. trachomatis antibody titer> 8 for seropositive patients; :$8 for seronegative patients. t p < 0.000001 (X 2 test); sensitivity: 74%; specificity: 92%; LR+: 9.1; X2 LR-: 0.3; OR: 31.5 (90% confidence interval: 8.3 to 138.5).
sitivity)/specificity). Calculation of LRs yields a score that allows categorization of test results: an LR+ between 2 and 5 indicates a fair clinical test, between 5 and 10 is good, and> 10 is excellent. An LR- between 0.5 and 0.2 indicates a fair clinical test, between 0.2 and 0.1 is good, and <0.1 is excellent (8). The odds ratio (OR) can be calculated from the LRs: OR = LR+ /LR- (12). The OR reflects the probability of a patient with an abnormal test to have tubal factor infertility. The OR equals 1 if the level of the test equals the prevalence of the disease. When the level exceeds the prevalence the OR is >1 (12). We calculated the LR+, LR-, and OR for C. trachomatis antibody testing and for HSG, both for our own study and for those literature reports allowing their calculation. The LRs and the OR for C. trachomatis antibody testing in our study were compared with the LRs and the OR for HSG, both from our study and from the literature.
RESULTS
In 60 of 211 patients studied (28.5%), the C. trachomatis antibody test was positive (anti-C. trachomatis IgG titer> 8). In 71 of 211 patients, a laparoscopy was performed (Table 1): 25 of the 28 seropositive patients (89%) had tuboperitoneal abnormalities. Of the 43 seronegative patients, only 9 (21 %) had adhesions or tubal obstruction. This difference is statistically highly significant (P < 0.000001; X 2 test). In 34 of 211 patients, an HSG as well as a laparoscopy was performed. In 24 of these 34 patients the results corresponded, but in 10 patients a discrepancy between HSG and laparoscopy was found (P = 0.06; Fisher's exact test) (Table 2). According to Bayes' theorem, the probability of tubal factor infertility for a patient in our popula-
Vol. 61, No.5, May 1994
tion is 88% if she has abnormal C. trachomatis antibody testing and 59% if she has an abnormal HSG. To obtain a prevalence independent measure of the reliability of C. trachomatis antibody testing, likelihood ratios were calculated. The LR+ of C. trachomatis antibody testing was 9.1, indicating a patient with tubal factor infertility to be 9.1 times more likely to have a positive test result (i.e., titer > 8) than a patient without tubal factor infertility. In comparison, the LR+ for HSG in the same group of patients was only 2.6. The LR - for C. trachomatis antibody testing was 0.3, indicating a patient with tubal factor infertility to be 0.3 times as likely to have a negative test (i.e., titer::;; 8) as a patient without the disease. The LR- for HSG was 0.5. The respective ORs were 31.5 for C. trachomatis antibody testing and 4.8 for HSG.
DISCUSSION
In the infertility patient, laparoscopy is the most accurate method to diagnose tuboperitoneal pathology but, for screening purposes, HSG is used widely. Hysterosalpingography is a disagreeable, annoying test to the patient. It may be painful and has a high false-positive and false-negative rate. The false-positive outcome (in most studies around 25 %) (13-23) is thought to result from tubal spasm, dissimilar tubal filling pressure, too high viscosity of the contrast medium, and faulty technique. False-negative findings have been reported in around 40% (13-23) and are considered to be due to peritubal adhesions not visualized on the delayed picture. We investigated the value of serum C. trachomatis antibody testing in screening for tubal factor infertility. With laparoscopy as a gold standard, we compared the predictive value of the C. trachomatis antibody titer with HSG. For this comparison we used LRs, prevalence independent test characteristics.
Table 2 Tuboperitoneal Abnormalities at Laparoscopy in Infertility Patients with Abnormal and Normal HSG
HSG
Abnormal laparoscopy
Normal laparoscopy
Abnormal Normal
7* 5
5 17
* p = 0.06 (Fisher's exact test); sensitivity: 58%; specificity: 77%; LR+: 2.6; LR-: 0.5; OR: 4.8 (90% confidence interval: 1.0 to 21.8).
Dabekausen et al.
Chlamydia antibody and tubal infertility
835
Table 3 LR+ and LR- of HSGs and Corresponding ORs in Predicting Tuboperitoneal Abnormalities as Calculated From Literature Reports Studies
LR+
LR-
OR*
Snowden et aI., 1984 (13) Moghissi and Kim, 1975 (14) Keirse and Vandervellen, 1973 (15) Gabos, 1976 (16) Swolin and Rosencrantz, 1972 (17) Present study Cumming and Taylor, 1980 (18) Ismajovich et aI., 1986 (19) Philipsen and Hansen, 1981 (20) World Health Organization, 1986 (21) Maathuis et aI., 1972 (22) Reshef et aI., 1989 (23)
6.1 4.9 4.6 2.9 2.7 2.6 2.3 2.0 2.0 1.8 1. 7 1.6
0.3 0.3 0.3 0.4 0.2 0.5 0.4 0.6 0.3 0.3 0.5 0.3
18.9 20.4 16.6 4.3 12.5 4.8 6.3 3.4 7.3 5.6 3.3 6.1
* Cumulative OR (90% confidence interval): 6.4 (5.3 to 7.9).
The LR+ of C. trachomatis antibody testing was 9.1; the LR+ ofthe HSG in this study was only 2.6. Theoretically, a poor performance of HSG in our clinic might be a reason for this disappointing test outcome. A review of the literature, however, showed LR+ of the HSG to range from 1.6 to 6.1 in those reports giving sufficient details for their calculation (Table 3). Even when the best LR+ of HSG in literature is compared with the LR+ ofthe C. trachomatis antibody test in our study, the latter reflects tuboperitoneal abnormalities better than HSG. The LR- of C. trachomatis antibody testing is 0.3 in our study and is comparable to the HSG studies with the best discriminatory value reported in the literature (Table 3). The group of patients who had HSG results available in our study was too small to allow for a reliable positioning of the new test, C. trachomatis antibody testing, in the diagnostic armamentarium to which the old test, HSG, belongs. Therefore, we estimated the cumulative OR and its confidence limits for HSG from a meta-analysis of those literature reports that provided sufficient details for their calculation. The cumulative OR for HSG was 6.4 with 90% confidence limits of 5.3 and 7.9. The OR of C. trachoma tis antibody testing differed significantly with an 0 R of 31.5 and 90% confidence limits of 8.3 and 138.5. Of the seronegative patients in our study, 9 of 43 (21 %) had tuboperitoneal abnormalities. These were due to previous pelvic surgery, endometriosis, or pelvic inflammatory disease caused by microorganisms other than C. trachomatis. In the present
836
Dabekausen et al.
study, the likelihood of disease increased with the height of the titer. All patients with a titer> 128 had tuboperitoneal disease. These results are in accordance with the findings by Minassian and Wu (24), who reported a correlation between the height of the titer and the severity of pelvic adhesions. Toye and co-workers (25) recently published a study of the association between antibody to the chlamydial heat-shock protein and tubal infertility. Their results suggest that chlamydial heat-shock protein determination allows further differentiation within the group of C. trachomatis-positive patients between those with and those without tubal damage. In conclusion, our study focused on serum anti -C. trachomatis IgG antibody screening in women presenting with infertility. C. trachomatis antibody testing by microimmunofluorescence is simple, inexpensive (Dfl 24 [$13] compared with Dfl 323 [$179] in The Netherlands) and causes minimal inconvenience to the patient. It is more likely than HSG to give abnormal results in patients with tubal factor infertility. C. trachomatis antibody testing deserves to become an integral component of the initial fertility work-up.
Chlamydia antibody and tubal infertility
REFERENCES 1. Hull MGR, Glazener CMA, Kelly NJ, Conway DI, Foster P A, Hinton RA, et al. Population study of causes, treatment, and outcome of infertility. Br Med J 1985;291:1693-7. 2. Arronet GH, Eduljee SY, O'Brien JR. A nine·year survey of fallopian tube dysfunction in human infertility. Fertil Steril 1969;20:903-18. 3. Paavonen J, Wolner-Hanssen P. Chlamydia trachomatis: a major threat to reproduction. Hum Reprod 1989;4:111-24. 4. Thejls H, Rahm VA, Rosen G, Gnarpe H. Correlation between chlamydia infection and clinical evaluation, vaginal wet smear, and cervical swab test in female adolescents. Am J Obstet Gynecol 1987;157:974-6. 5. Schoenwald E, Schmidt BL, Steinmetz G, Hosmann J, Pohla·Gubo G, Luger A, et al. Diagnosis of Chlamydia trachomatis infections. Culture versus serology. Eur J Epide· miol 1988;4:75-82. 6. Puolakkainen M, Vesterinen E, Purola E, Saikku P, Paavonen J. Persistence of Chlamydial antibodies after pelvic inflammatory disease. J Clin Microbiol 1986;23:924-8. 7. 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. 8. Collins JA. Male infertility: the interpretation of the diagnostic assessment. In: Mishell DR, Paulsen CA, Lobo RA, editors. Yearbook of infertility 1989. Chicago: Year Book Medical Publishers, 1989:45-75. 9. Treharne JD, Darougar S, Jones BR. Modification of the micro-immunofluorescence test to provide a routine sero-
Fertility and Sterility
diagnostic test for chlamydial infection. J Clin Pathol 1977;30:510-7. 10. Barnes RC. Laboratory diagnosis of human chlamydial infections. Clin Microbiol Rev 1989;2:119-37. 11. McNeil BJ, Hanley JA. Statistical approaches to the analysis of receiver operating characteristics. Med Decis Making 1984;4:137-50. 12. Kraemer HC. Evaluating medical tests, objective and quantitative guidelines. Newbury Park: Sage Publications, 1992:103-13. 13. Snowden EU, Jarrett JC II, Dawood MY. Comparison of diagnostic accuracy of laparoscopy, hysteroscopy, and hysterosalpingography in evaluation of female infertility. Fertil Steril1984;41:709-13. 14. Moghissi KS, Sim GS. Correlation between hysterosalpingography and pelvic endoscopy for the evaluation of tubal factor. Fertil Steril 1975;26:1178-81. 15. Keirse MJ CN, Vandervellen R. A comparison of hysterosalpingography and laparoscopy in the investigation of infertility. Obstet Gynecol 1973;41:685-8. 16. Gabos P. A comparison of hysterosalpingography andendoscopy in evaluation of tubal function in infertile women. Fertil Steril 1976;27:238-42. 17. Swolin K, Rosencrantz M. Laparoscopy vs. hysterosalpingography in sterility investigations: a comparative study. Fertil Steril 1972;23:270-3. 18. Cumming DC, Taylor PJ. Combined laparoscopy and hys-
Vol. 61, No.5, May 1994
19.
20.
21.
22.
23.
24.
25.
teroscopy in the investigation of the ovulatory infertile female. Fertil SteriI1980;33:475-8. Ismajovich B, Wexler S, Golan A, Langer L, David MP. The accuracy of hysterosalpingography versus laparoscopy in evaluation of infertile women. Int J Gynaecol Obstet 1986;24:9-12. Philipsen T, Hansen BB. Comparative study of hysterosalpingography and laparoscopy in infertile patients. Acta Obstet Gynecol Scand 1981;60:149-51. WHO study. Comparative trial of tubal insufflation, hysterosalpingography, and laparoscopy with dye hydrotubation for assessment of tubal patency. Fertil Steril1986;46:11017. Maathuis JB, Horbach JGM, van Hall EV. A comparison of the results of hysterosalpingography and laparoscopy in the diagnosis of fallopian tube dysfunction. Fertil Steril 1972;23:428-31. Reshef E, Daniel WW, Foster JC, Bradley EL, Blackwell RE, Younger JB. Comparison between I-hour and 24-hour follow-up radiographs in hysterosalpingography using oil based contrast media. Fertil Steril 1989;52:753-5. Minassian SS, Wu CH. Chlamydia antibody by enzymelinked immunosorbent assay and associated severity of tubal factor infertility. Fertil Steril1992;58:1245-7. Toye B, Caferriere 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.
Dabekausen et al.
Chlamydia antibody and tubal infertility
837