Department of Obstetrics and Gynecology, Kyoto Prefectural University of Medicine, Graduate School of Medical Science, Kyoto 602-8566, Japan
1 To whom correspondence should be addressed at: Jo Kitawaki, M.D., Ph.D., Department of Obstetrics and Gynecology, Kyoto Prefectural University of Medicine, 465 Kajii-cho, Kamigyo-ku, Kyoto 602-8566, Japan. Fax: +81-75-212-1265; Email: kitawaki{at}koto.kpu-m.ac.jp
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: adenomyosis/CA-125/diagnosis/endometriosis/leiomyoma/ovarian endometriomas
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Serologic testing for CA-125 has been widely used for detection of endometriosis and monitoring of progressive disease (Barbieri et al., 1986; Pittaway and Fayez, 1986
; Fedele et al., 1989
; Koninckx et al., 1996
; Medl et al., 1997
; Cheng et al., 2002
). Although the standard cutoff value of 35 U/mL was initially set to detect epithelial ovarian cancer (Bast et al., 1983
), a decisive cutoff value for screening of endometriosis has not been set because serum level does not necessarily correlate with severity of disease. Indeed, serum levels in women with mild endometriosis are often lower than those in women without endometriosis. A meta-analysis based on 23 articles showed a limited diagnostic performance of serum CA-125 in detecting endometriosis (Mol et al., 1998
). However, only a limited number of the included studies had a sufficient number of patients to set a cutoff value (Barbieri et al., 1986
; Pittaway and Fayez, 1986
; Fedele et al., 1989
; Koninckx et al., 1996
; Medl et al., 1997
). These larger studies confirmed an elevation of serum CA-125 level in endometriosis, including both mild and severe stages, but not in mild endometriosis without ovarian involvement.
In order to evaluate the diagnostic value of CA-125 assays for possible cases of mild endometriosis, we measured serum CA-125 levels for 775 consecutive women undergoing laparotomy or laparoscopy; we analyzed results separately for women with mild disease without ovarian endometriomas and women with severe disease with endometriomas. We also did separate analyses based on association of endometriosis with adenomyosis and leiomyomas, both of which might affect serum CA-125 level.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Blood samples were drawn before surgery on days other than those during menstruation. Serum concentrations of CA-125 were measured by an immunoradiometric assay kit using the specific monoclonal antibody M11 (Centocor, Malvern, PA, USA) and were expressed in arbitrary units based on a primary standard. The intra- and inter-assay coefficients of variation were 3.4% at 42.6±1.4 U/mL and 5.3% at 40.7±2.1 U/mL, respectively.
Statistical analyses were performed with Statflex version 5.0 software (Artech, Osaka, Japan). The difference in serum CA-125 levels between disease groups was analyzed by the Kruskal-Wallis test followed by nonparametric Dunn's test. A receiver operating characteristic curve (ROC) analysis was carried out by plotting the false-positive rate (1specificity) versus the true-positive rate (sensitivity) for increasing cutoff values at an interval of 1 U/mL. Positive predictive values (PPV), negative predictive values (NPV), accuracy, and diagnostic value (defined as specificityxsensitivity) were also calculated. The difference between the areas under the ROC curves was analyzed by Harley's test. A P-value of <0.05 was considered significant.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
The ROC curve analysis revealed that the areas under the curves for both groups with endometriomas (e.g. E and E/A/L) were greater than 0.9, indicating a good diagnostic performance (Figure 2, Table II). However, the area under the curve for E without endometriomas was 0.788, significantly smaller than that for E with endometriomas (P<0.05). The area for A/L was significantly smaller than those for E with endometriomas and E/A/L with endometriomas (P<0.001) (Figure 2, Table II).
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
In practice, severe endometriosis can be diagnosed with high accuracy by pelvic examination. Transvaginal ultrasonography is routinely used, and it detects severe endometriosis associated with ovarian endometriomas with both high sensitivity and specificity (Guerriero et al., 1998; Patel et al., 1999
; Dessole et al., 2003
). Magnetic resonance imaging also detects endometriomas with high accuracy (Takahashi et al., 1994
; Bis et al., 1997
; Kinkel et al., 1999
; Stratton et al., 2003
). Adenomyosis and leiomyomas can be detected with high accuracy by pelvic examination and imaging tools (Eskenazi et al., 2001
). However, these imaging tools are still not sensitive enough to detect mild endometriosis associated with peritoneal implants but not with endometriomas. Given such a situation, a serological test is most useful if it is applicable to detect endometriosis without endometriomas. CA-125 in the systemic circulation originates mostly from ectopic endometriotic tissues and, to a lesser extent, from eutopic endometrium and ovary. However, immunohistological staining intensity does not correlate with the serum level (Toki et al., 2000
). Although a number of studies have investigated the usefulness of serum testing for CA-125, most of these studies analyzed diagnostic performance combining mild and severe endometriosis (Barbieri et al., 1986
; Pittaway and Fayez, 1986
; Fedele et al., 1989
; Koninckx et al., 1996
; Medl et al., 1997
; Cheng et al., 2002
). There has been no study discriminating endometriosis with or without endometriomas. It can be estimated that the studies would have obtained a lower diagnostic value if the cases had been limited to mild endometriosis. Mol et al. (1998)
reported a meta-analysis based on 23 articles; they analyzed the diagnostic performance of serum CA-125 measurement for mild plus severe endometriosis and severe endometriosis only. As expected, they obtained a better diagnostic value for severe endometriosis than combined with mild endometriosis. In diagnosing all grades of endometriosis at a specificity of 90%, sensitivity was only 28%. If sensitivity were increased to 50%, specificity dropped to 72%. In diagnosis of severe endometriosis with a specificity of 89%, sensitivity increased to 47%, and if sensitivity were increased to 60%, specificity increased to 81%.
In the present study, the mean serum CA-125 level increased when cases were associated with ovarian endometriomas. Serum CA-125 level increased with grade of endometriosis, consistent with findings from most previous studies (Barbieri et al., 1986; Pittaway and Fayez, 1986
; Fedele et al., 1989
; Koninckx et al., 1996
; Medl et al., 1997
; Cheng et al., 2002
). As expected, a good diagnostic performance was obtained for endometriosis with endometriomas. However, it should be noted that as many as 10.6% of cases of E with endometriomas and 15.6% of cases of E/A/L with endometriomas had serum CA-125 levels less than 20 U/mL. In contrast, the area under the ROC curve for E without endometriomas was 0.788, indicating that the diagnostic accuracy for endometriosis without endometriomas was naturally less than that for endometriosis with endometriomas. The present study provided sufficiently high specificity but insufficiently low sensitivity compared with reports in other studies (Barbieri et al., 1986
; Pittaway and Fayez, 1986
; Fedele et al., 1989
; Koninckx et al., 1996
; Medl et al., 1997
; Mol et al., 1998
; Cheng et al., 2002
). Accordingly, NPV was always lower than PPV.
Our population was one of the largest series comprising consecutive patients among the studies that have analyzed CA-125 levels and diagnosis of endometriosis. This might have resulted in reducing selection bias and minimizing errors in statistical values. Because PPV and NPV are influenced by the frequency of disease, this type of study offers an advantage for assessing diagnostic performance.
Given the results, although the cutoff value 20 U/mL showed the best diagnostic performance, we propose a combined use of two cutoff values, 20 and 30 U/mL, rather than a single cutoff value in diagnosis of endometriosis without endometriomas. When focused on group E without endometriomas, if a serum CA-125 level equals or is higher than 30 U/mL, the probability of endometriosis is 92.9% (PPV). However, even if a level is less than 30 U/mL, the possibility of endometriosis cannot be neglected. By contrast, if a CA-125 level is less than 20 U/mL, endometriosis can be ruled out with a probability of 78.0% (NPV). If CA-125 level equals or is higher than 20 U/mL and less than 30 U/mL, diagnostic performance is limited. This setting for cutoff values is clearly superior to the single empirical cutoff value of 35 U/mL. However, the accuracy of using only CA-125 testing for diagnosis is still limited. In clinical practice, 93% of women with a serum CA-125 level of 30 U/mL or higher who have had the diagnoses of adenomyosis, leiomyomas, ovarian tumors, pelvic inflammation, and/or pregnancy ruled out will be shown to have endometriosis on further diagnostic testing. At the lower end of possible cutoff levels, 22% of women with a serum CA-125 level lower than 20 U/mL in the same diagnostic situation will still be shown to have endometriosis.
In conclusion, research to date indicates that serum CA-125 testing appears to be the least invasive, least expensive, simplest, and most rapid method to screen for the diagnosis of endometriosis. This test can be positioned for initial screening together with transvaginal ultrasonography. In the diagnosis of endometriosis without endometriomas, combined use of two cutoff values for CA-125, 20 and 30 U/mL, provides improved diagnostic performance. On the other hand, the present results also show the limit of serum CA-125 testing. Several less invasive tests have been postulated {e.g. testing aromatase (Kitawaki et al., 1999) and leukocyte subsets (Gagne et al., 2003
)} for use in biopsied specimens of eutopic endometrium, and these techniques look promising. These moderately invasive techniques can be positioned as secondary tests before laparoscopy. Such systematic testing will provide more accurate diagnosis of endometriosis through a series of increasingly invasive tests.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Barbieri RL, Niloff JM, Bast RC Jr, Scaetzl E, Kistner RW and Knapp RC (1986) Elevated serum concentrations of CA-125 in patients with advanced endometriosis. Fertil Steril 45, 630634.[ISI][Medline]
Bast RC Jr, Klug TL, St. John E, Jenison E, Niloff JM, Lazarus H, Berkowitz RS, Leavitt T, Griffiths CT, Parker L et al. (1983) A radioimmunoassay using a monoclonal antibody to monitor the course of epithelial ovarian cancer. N Engl J Med 309, 883887.[Abstract]
Bedaiwy MA, Falcone T, Sharma RK, Goldberg JM, Attaran M, Nelson DR and Agarwal A (2002) Prediction of endometriosis with serum and peritoneal fluid markers: a prospective controlled trial. Hum Reprod 17, 426431.
Bis KG, Vrachliotis TG, Agrawal R, Shetty AN, Maximovich A and Hricak H (1997) Pelvic endometriosis: MR imaging spectrum with laparoscopic correlation and diagnostic pitfalls. Radiographics 17, 639655.[Abstract]
Cheng YM, Wang ST and Chou CY (2002) Serum CA-125 in preoperative patients at high risk for endometriosis. Obstet Gynecol 99, 375380.
Dessole S, Farina M, Rubattu G, Cosmi E, Ambrosini G and Nardelli GB (2003) Sonovaginography is a new technique for assessing rectovaginal endometriosis. Fertil Steril 79, 10231027.[CrossRef][ISI][Medline]
Eskenazi B, Warner M, Bonsignore L, Olive D, Samuels S and Vercellini P (2001) Validation study of nonsurgical diagnosis of endometriosis. Fertil Steril 76, 929935.[CrossRef][ISI][Medline]
Fedele L, Arcaini L, Vercellini P, Marchini M, Baglioni A and Bianchi S (1989) Serum Ca-125 concentrations in endometriosis. Acta Eur Fertil 20, 137139.[Medline]
Gagne D, Rivard M, Page M, Lepine M, Platon C, Shazand K, Hugo P and Gosselin D (2003) Development of a nonsurgical diagnostic tool for endometriosis based on the detection of endometrial leukocyte subsets and serum CA-125 levels. Fertil Steril 80, 876885.[CrossRef][ISI][Medline]
Guerriero S, Ajossa S, Mais V, Risalvato A, Lai MP and Melis GB (1988) The diagnosis of endometriomas using colour Doppler energy imaging. Hum Reprod 13, 16911695.[CrossRef]
Kinkel K, Chapron C, Balleyguier C, Fritel X, Dubuisson JB and Moreau JF (1999) Magnetic resonance imaging characteristics of deep endometriosis. Hum Reprod 14, 10801086.
Kitawaki J, Kusuki I, Koshiba H, Tsukamoto K, Fushiki S and Honjo H (1999) Detection of aromatase cytochrome P-450 in endometrial biopsy specimens as a diagnostic test for endometriosis. Fertil Steril 72, 11001106.[CrossRef][ISI][Medline]
Koninckx PR, Meuleman C, Oosterlynck D and Cornillie FJ (1996) Diagnosis of deep endometriosis by clinical examination during menstruation and plasma CA-125 concentration. Fertil Steril 65, 280287.[ISI][Medline]
Medl M, Ogris E, Peters-Engl C, Mierau M, Buxbaum P and Leodolter S (1997) Serum levels of the tumour-associated trypsin inhibitor in patients with endometriosis. Br J Obstet Gynaecol 104, 7881.[ISI][Medline]
Mol BW, Bayram N, Lijmer JG, Wiegerinck MA, Bongers MY, van der Veen F and Bossuyt PM (1998) The performance of CA-125 measurement in the detection of endometriosis: a meta-analysis. Fertil Steril 70, 11011108.[CrossRef][ISI][Medline]
Patel MD, Feldstein VA, Chen DC, Lipson SD and Filly RA (1999) Endometriomas: diagnostic performance of US. Radiology 210, 739745.
Pittaway DE and Fayez JA (1986) The use of CA-125 in the diagnosis and management of endometriosis. Fertil Steril 46, 790795.[ISI][Medline]
Stratton P, Winkel C, Premkumar A, Chow C, Wilson J, Hearns-Stokes R, Heo S, Merino M and Nieman LK (2003) Diagnostic accuracy of laparoscopy, magnetic resonance imaging, and histopathologic examination for the detection of endometriosis. Fertil Steril 79, 10781085.[CrossRef][ISI][Medline]
Takahashi K, Okada S, Ozaki T, Kitao M and Sugimura K (1994) Diagnosis of pelvic endometriosis by magnetic resonance imaging using "fat-saturation" technique. Fertil Steril 62, 973977.[ISI][Medline]
Toki T, Kubota J, Lu X and Nakayama K (2000) Immunohistochemical analysis of CA125, CA19-9, and Ki-67 in stage III or IV endometriosis: positive correlation between serum CA125 level and endometriotic epithelial cell proliferation. Acta Obstet Gynecol Scand 79, 771776.[CrossRef][ISI][Medline]
Submitted on November 5, 2004; resubmitted on February 18, 2005; accepted on February 28, 2005.
|