Human Reproduction vol.12 no.ll pp.2523-2527, 1997
The use of biochemical markers in the diagnosis of
pelvic endometriosis
Mauricio S.Abrao1'3, Sergio Podgaec1, Braz
Martorelli Filho1, Laudelino O.Ramos1, Jose
Aristodemo Pinotti1 and Ricardo M.de Oliveira2
Departments of 'Obstetrics and Gynecology and 2Clinical
Pathology, Sao Paulo University School of Medicine, Sao Paulo,
SP, Brazil
3
To whom correspondence should be addressed at: Rua Antonio
Jose de Almeida, 174 Sao Paulo, SP 04720-060, Brazil
The aim of this study was to evaluate CA 125 II, Creactive protein (CRP) and serum amyloid A (SAA) and
anticardiolipin antibody (aCL) concentrations for the diagnosis of pelvic endometriosis. The study population consisted of 15 women without endometriosis, as confirmed
by laparoscopy (group A), and 35 patients with pelvic
endometriosis diagnosed by laparoscopy or laparotomy
(group B). Group B patients were divided into those at
stages I and II of the disease (BI/II) and those at stages III
and IV (BIII/IV). Blood samples were obtained twice during
the menstrual cycle: on day 1, 2 or 3 of the cycle and on
day 8, 9 or 10 of the cycle. CA 125 II and CRP concentrations were higher in group III/IV patients compared with
healthy controls, mainly during the first 3 days of the
menstrual cycle; SAA concentrations were also higher in
this group of patients compared with healthy controls,
but only during the first 3 days of the menstrual cycle.
Immunoglobulin (Ig) M aCL concentrations were higher
in all patients with endometriosis compared with healthy
controls, mainly during the first 3 days of the menstrual
cycle. It is concluded that these determinations may contribute to the diagnosis and the indication of treatment for
pelvic endometriosis. Determination of CA 125 II concentrations at the beginning of the menstrual cycle may aid
the diagnosis of stage III and IV endometriosis. IgM aCL
appears to be associated with the presence of all stages of the
disease, while SAA values are elevated in severe situations.
Measurement of these molecules may therefore provide a
valuable tool in the diagnosis and management of endometriosis.
Key words: C-reactive protein/CA 125/cardiolipins/endometriosis/serum amyloid protein A
Introduction
Endometriosis has been the subject of many published reports
in the medical literature and is considered to be an enigmatic
disease of uncertain aetiology and variable behaviour. The
condition is responsible for major problems in women in the
reproductive years, among them pelvic pain and infertility.
European Society for Human Reproduction and Embryology
Many studies and theories also exist about the aetiopathogenesis and diagnosis of the disease. Laparoscopy still represents
the most accurate investigative method (Izzo et al, 1984;
Evers et al, 1995; Maclavert and Shaw, 1995). Considerable
efforts are currently being devoted to the identification of
possible markers of the disease to make its diagnosis less
invasive and more accessible (Evers et al, 1995). The first
and most frequently used marker is CA 125, and its application
was first reported by Bast et al in 1983. Niloff et al (1984)
reported increased serum CA 125 concentrations in patients
with endometriosis. Several other studies followed, showing
varying specificity and sensitivity rates with different cutoff points (Pittaway and Fayez, 1986; Fedele et al, 1988;
Barbieri, 1990).
To improve its sensitivity, a method of measuring CA 125
II, based on the combination of two monoclonal antibodies
with the ability to bind to the antigen at different sites, has
been developed (O'Brien et al, 1991). Hornstein et al (1994)
compared CA 125 with CA 125 II in patients with endometriosis. They reported an improved diagnostic specificity and
sensitivity with the use of CA 125 II in the various stages of
the disease.
Anticardiolipin antibodies (aCL) are antiphospholipid antibodies detected by assays based on the use of cardiolipin
(diphosphatidyl glycerol) in the solid phase. Kennedy et al.
(1989) reported a significant increase in anticardiolipin
immunoglobulin (Ig) G antibody in patients with endometriosis
and systemic lupus erythematosus compared with those found
in healthy women. Kilpatrick et al. (1991) concluded that only
in cases of endometriosis associated with infertility was there
a slight increase in the concentrations of these antibodies,
although the difference was not significant.
Proteins of the acute inflammatory phase, such as C-reactive
protein (CRP) and serum amyloid A (SAA), have been
described as indirect markers of the intensity of diseases that
involve inflammatory processes. The increase in CRP in the
acute inflammatory response seems to be directly proportional
to the amount of tissue damage, and the major use of this
marker has been in rheumatic diseases such as rheumatoid
arthritis and systemic lupus erythematosus. SAA is measured
by an immunoenzymatic reaction, and its clinical application
currently ranges from systemic amyloidosis to other inflammatory states.
The discordance in the literature about the use of CA 125
in patients with endometriosis, the absence of studies on the
best time to measure CA 125 II, and the absence of reports
on the participation of proteins of the acute inflammatory
phase and on the optimum time during the menstrual cycle at
2523
M.Abrao et al.
which antibodies of the anticardiolipin type should be measured
led us to undertake this study.
To assess the extent of association between each characteristic
(individually or as a whole) and the presence or absence of disease,
a logistic regression model was used (Kleinbaum et al, 1982).
Materials and methods
Cases
A total of 50 women who were consulted between January 1993 and
June 1995 were selected and divided into two groups. The first group
(group A; control) consisted of 15 women with no pelvic pain and
with a history of bilateral tube ligation performed elsewhere who had
approached our institution to reverse the ligation in an attempt
to become pregnant. These patients were submitted to diagnostic
laparoscopy to rule out other infertility factors before the execution
of tube reanastomosis, when the absence of pelvic endometriosis was
confirmed. The second group (group B) consisted of 35 women
with pelvic endometriosis which was confirmed by laparoscopy or
laparotomy.
The inclusion criteria for the two groups were as follows: Group
A (control): absence of current or previous endometriosis; absence
of pelvic pain in the form of dysmenorrhoea, chronic pelvic pain or
dyspareunia; absence of diseases of the uterus, uterine tubes or ovaries
confirmed by transvaginal pelvic ultrasound and by laparoscopy; age
ranging from 20 to 40 years. Group B (endometriosis): absence of
previous clinical or surgical treatment of endometriosis; absence of
other diseases of the uterus, tubes or ovaries; a diagnosis of pelvic
endometriosis confirmed by biopsy; age ranging from 20 to 40 years.
Patient age ranged from 26 to 38 years in group A (mean 34.01
± 0.98 years) and from 20 to 40 years in group B (mean 30.63 ±
1.46 years).
Methods
Study design
All women in groups A and B were submitted to serum analysis of
CA 125 II, CRP, SAA protein and aCL. The samples were collected
at two different times during the menstrual cycle, i.e. on day 1, 2 or
3 of the cycle (time 1) and on day 8, 9 or 10 of the cycle (time 2).
In group A, the determinations were made 20-90 days after laparoscopy to eliminate interference from the diagnostic procedure with the
laboratory data; in group B, the determinations were made during the
3 months preceding laparoscopy or laparotomy.
In group B, endometriosis was diagnosed by video laparoscopy in
29 cases and by laparotomy in six. The disease was classified into
four stages according to the revised guidelines of the American
Fertility Society (1985). To guarantee a reasonable number of patients
in each subgroup without affecting the results, stages I and II,
and stages III and IV, were grouped together (BI/II and BIII/IV
respectively).
For a more complete analysis of the data, in addition to an analysis
of each phase of the menstrual cycle, we also calculated for each
characteristic under study the ratio of the values obtained at time 1
and time 2 and the difference between these values.
Laboratory tests
The collected blood samples were centrifuged at 1200 r.p.m. for 4
min and the supernatant was stored at -20°C until the examination.
CA 125 II was then analysed by chemiluminescence, and CRP, SAA
protein and aCL were analysed by an enzyme immunoassay (Loose
et al, 1993; de Oliveira et al, 1994).
Statistical analysis
Data concerning categorized characteristics were analysed by Fisher's
exact (generalized) test and the likelihood ratio test (%2) (Agresti,
1990). Data concerning continuous characteristics were analysed by
an analysis of variance (ANOVA; Neter and Wasserman, 1974).
2524
Results
The mean values obtained at the two points during the
menstrual cycle (time 1 and time 2) and the respective ratios
and differences for each variable studied are presented in
Table I.
An ANOVA for serum CA 125 II at the two points during
the menstrual cycle and for the differences between the values
obtained at the two points was carried out to compare the
control group with groups BI/II and BIII/IV (Table II).
We further compared the characteristics of the patients at
stages I/II and III/IV of the disease with those for patients
without the disease. Except for SAA at time 2 and for
IgG aCL at both observation points, the distributions of the
remaining characteristics for group BIII/IV patients were
significantly different from the distributions for patients without
the disease (P < 0.01) both at time 1 and at time 2. In contrast,
when BI/II patients were compared with controls, no significant
differences were observed (P > 0.05), except for IgM aCL
distribution at time 1 (P = 0.015).
To assess the extent of association between each characteristic (individually and as a whole) and the presence or
absence of endometriosis, we used logistic regression models
(Kleinbaum et al, 1982). These can be used to estimate the
expected proportion of patients with the disease on the basis
of the values of the characteristics in question. For this
purpose we only considered the determinations made at time
1. Furthermore, to reduce the variability of CA 125 II and
thus obtain more stable estimates, we used the square root of
its values in the models employed.
When the characteristics under study were examined individually, the model based on CA 125 II was the one that best
fitted the data. In addition, the models based exclusively on
CRP and on IgM aCL also presented significant results.
However, in the specific case of CRP, the discrimination
between patients with the disease and controls was significant
only for patients with CRP values > 3 .
In the presence of CA 125 II, none of the other characteristics
investigated made a statistically significant contribution. The
reason is that CA 125 II better explains the variability of the
data. Table I shows that the control group had a mean CA 125
II value of 14.61 ± 3.06 at time 1 (range 0.50^48.60), while
the group with endometriosis had a mean CA 125 II value of
148.36 ± 39.19 at time 1 (range 27.40-893.00).
The model obtained using a relevant statistical formula may
be presented as follows:
odds to have the disease = 1/[1 + exp (-y)],
where y = -10.300 + 1.944 V CA 125 (time 1).
According to this model, the odds of a patient having the
disease compared with the odds of not having it increase by
6.99 (confidence interval 1.61-30.20) per unit increase in the
square root of CA 125 II at time 1. Furthermore, we may
obtain estimates of the expected proportion of patients with
Markers in the diagnosis of endometnosis
Table I. Mean values and SD of serum CA 125 II, C-reactive protein (CRP), serum amyloid A (SAA) protein, immunoglobulin (Ig) G and IgM
anticardiolipin antibodies (aCL) at times 1 and 2, with their respective ratios and differences between the two time points of the menstrual cycle for the
control group (A) and patients with endometriosis (BI/II and BIII/IV)
Group
Time 1
Time 2
Ratio: time I/time 2
Difference: time 1 - time 2
A
BI/II
BIII/IV
A
BI/II
BIII/IV
A
BI/II
BIII/IV
A
BI/II
BIII/IV
n
15
20
15
15
20
15
15
20
15
15
20
15
CA 125 (IU/ml)
CRP (ng/ml)
SAA (Hg/ml)
Mean
SD
Mean SD
Mean
14.61
73.37
248.35
9.69
36.49
138.17
1.40
2.13
2.02
4.92
36.89
110.19
3.06
11.26
58.72
1.62
4.51
32.57
0.15
0.20
0.30
1.72
8.77
46.58
1.58
5.06
13.15
1.53
2.89
3.52
1.46
2.88
3.58
0.05
2.17
9.63
7.25
8.04
38.82
4.79
4.17
6.85
1.91
2.26
6.98
2.46
3.87
31.97
0.31
1.32
2.55
0.33
0.59
0.41
0.31
0.63
0.60
0.38
1.16
2.34
IgG aCL (GPL)
IgM aCL (MPL)
SD
Mean
SD
Mean
SD
2.04
1.93
16.99
1.58
0.85
1.92
0.65
0.52
1.58
1.97
1.61
16.43
4.26
11.54
16.24
3.73
7.53
8.41
1.34
8.74
34.60
0.53
4.01
7.83
0.48
3.86
6.60
0.42
2.09
2.84
0.19
5.37
33.24
0.58
2.06
4.59
6.11
15.59
25.33
5.71
10.05
14.52
1.07
8.43
2.03
0.40
5.55
10.81
1.27
3.03
5.67
0.74
2.07
2.22
0.17
6.51
0.54
0.91
1.59
4.38
GPL = units of anticardiolipin antibodies of the IgG type; MPL = units of anticardiolipin antibodies of the Igm type.
Table II. Analysis of variance for serum CA 125 II at the two time points of the menstrual cycle and the
difference in values between these two time points for the control group (A) and patients with endometriosis
(BI/II and BIII/IV)
Group
Time 2
Time 1
A
BI/II
BIII/IV
Difference: time 1time 2
Mean
(IU/ml)
Significance
level
Mean
(IU/ml)
Significance
level
Mean
(IU/ml)
Significance
level
14.61
73.37
248.35
P > 0.05
P < 0.05
9.69
36.48
138.17
P > 0.05
P < 0.05
4.92
36.89
110.19
P > 0.05
P < 0.05
The significance levels corresponding to an analysis of variance for simultaneous comparison of the three
groups were P = 0.0001, P = 0.0001 and P = 0.0197 respectively.
Table III. Estimated proportion of patients with endometriosis for different
CA 125 II values at time 1 during the menstrual cycle
CA125 II at time 1 (IU/ml)
Estimated proportion of patients
with endometriosis (%)
16
33
100
7
70
100
SE
(%)
the disease for different CA 125 II values at time 1. Some
examples are presented in Table III.
Excluding CA 125 II, an adequate model is that involving
CRP and IgM aCL. This model with only two CRP categories
(above and below 3) may be represented as follows:
odds to have the disease = 1/[1 + exp (-y)],
where y = -1.712 + 2.957 Category (CRP) + 2.715 Category
(IgM aCL).
In the above equation we attributed a value of 0 to the CRP
category when the CRP concentration was =^3 Jig/ml, and a
value of 1 when the CRP concentration was > 3 |i,g/ml. For
IgM aCL we attributed a value of 0 when the IgM aCL
concentration was =sl0 MPL and a value of 1 when the IgM
aCL concentration was >10 MPL.
According to this model, the odds of a woman having the
disease compared with the odds of a woman not having it is
0.18 (confidence interval 0.04-0.86) in the presence of CRP
concentrations =s3 |ig/ml and IgM aCL concentrations ^10
MPL. These odds are multiplied by 19.20 (confidence interval
0.04-0.86) when CRP concentrations are >3 |ug/ml and
by 15.20 (confidence interval 2.21-103.00) when IgM aCL
concentrations are > 10 MPL at time 1 (Table IV).
Discussion
One of the problems concerning endometriosis today is its
high incidence of 5-15% in women during the reproductive
years (Barbieri, 1990; West, 1990), and the difficulty in
diagnosing it (Galle, 1989; Berger, 1995). There is currently
much debate about the advantage of treating endometriosis at
its early stages (Maclavert and Shaw, 1995). Thus, many
studies of the disease have been based on the possible use of
new markers or recent imaging techniques (Bonilla-Musoles
and Martinez-Molina, 1995; Evers et ai, 1995; Maclavert and
Shaw, 1995).
The major objective of our study was to evaluate the
behaviour of markers in women with endometriosis and in
healthy women. The major marker cited in the literature has
been CA 125, which also increases in situations of inflammation, such pelvic inflammatory disease. The decision to collect
2525
M.Abrao et al.
Table IV. Estimated proportion of patients with endometriosis according to the determination of Creactive protein (CRP) and immunoglobulin (Ig) M anticardiolipin antibodies (aCL)
Proportion of patients
Category CRP
(|J.g/ml)
IgM aCL
(MPL)
Total no. of
patients
No. of patients
with endometriosis
Observed
Estimated
SE (%)
^3
=£3
>3
>3
=£10
>10
=£10
>10
11
10
12
17
2
7
9
17
18
70
75
100
15
73
78
98
10
13
11
2
blood at two different time points during the menstrual
cycle was made from studies showing oscillations in its
concentrations during the menstrual cycle (Pittaway and Fayez,
1987). The progress of endometriosis is under the influence
of oestrogens, whose levels are more pronounced during week
2 of the menstrual cycle. During menstruation inflammation
occurs around the focus of the disease, which may explain the
increase in CA 125 concentrations. Thus it is possible that
proteins of the acute inflammatory phase, such as CRP and
SAA or aCL, may also vary in a similar manner.
The measurements made at time 1 (day 1, 2 or 3 of the
menstrual cycle) differentiated the three groups in terms of all
characteristics more than those made at time 2 (day 8, 9 or 10
of the menstrual cycle), demonstrating that the increase in
values obtained was clearly greater during the first 3 days of
menstruation (Table I). This may be due to the association
of the disease with inflammatory phenomena, as discussed
previously.
Considering CA 125 separately, the greater usefulness of its
determination at time 1 is in sharp contrast to the results of
Pittaway and Fayez (1987), who reported an increase in CA
125 values during menstruation in patients with and without
endometriosis. This may be explained in two ways. First, these
investigators were not measuring CA 125 II. Second, the more
recent reports of new laparoscopic images compatible with
endometriosis, many of which appeared at the end of the
1980s after the study by Pittaway and Fayez (1987), have
perceptibly changed the approach to the disease (Jansen and
Russel, 1986; Stripling et al, 1988).
In patients with early stages of the disease (BI/II group),
CA 125 II alone was not statistically significant for the
diagnosis of endometriosis. Its use in diagnosis requires other
information such as clinical and imaging data. It should be
pointed out that CA 125 is useful to monitor patients receiving
treatment for endometriosis. If clinical treatment with antioestrogen drugs causes amenorrhoea, a previous measurement
during the mid-follicular phase (time 2) may be useful for
comparison with CA 125 values obtained at follow-up.
The acute phase proteins CRP and SAA have been studied
extensively over the last few years. Recently more sensitive
techniques have been developed for their estimation (de
Oliveira et al, 1994). However, no reports are available about
their role in endometriosis. The association of these proteins
with interleukins 1 and 6 and with tumour necrosis factor, also
involved in endometriosis (Gleicher et al, 1993; Dmowski,
1995), not only indicates their potential use in the diagnosis
2526
of the disease, but also suggests that future studies should be
conducted on the treatment of endometriosis using drugs that
act on the release of these cytokines. This potential therapy
could be monitored by measurement of these serum acute
phase proteins.
The analysis of these proteins at time 1 also permitted better
differentiation of the groups. Table I shows that the CRP
values obtained for the control group were similar at the two
different time points during the menstrual cycle. CRP values
were increased at time 1 in both the BI/II and BIII/IV groups,
with a clear increase in the latter. SAA measurement provided
additional information (Table I). SAA values were not significantly increased in group BI/II, in contrast to group BIII/IV.
The significant increase in group BIII/IV at time 1 indicates
that SAA values >10 |lg/ml may suggest the presence of
advanced disease.
There are discrepancies in the literature about the participation of aCL in endometriosis or in spontaneous abortion
(Kilpatrick et al, 1991; Abrao et al, 1992; Gleicher et al,
1993; Dmowski, 1995). Odukoya et al. (1996) studied the
sCD 23 protein, the low affinity receptor for the Fc portion of
IgE which is expressed on B cells, monocytes/macrophages
and thymic epithelial cells. They concluded that in patients
with endometriosis the concentration of this protein is higher
in the peritoneal fluid than in controls, suggesting the presence
of B cell activation in patients with this disease. They found
no association between the peritoneal fluid concentration of
this protein and the phase of the menstrual cycle, but only one
sample per patient was obtained by laparoscopy.
As was also the case for CA 125 II, CRP and SAA, the
aCL determinations made at time 1 differentiated the three
groups more than those made at time 2 (Table I). This
information suggests the use of IgM aCL in the diagnosis of
endometriosis. It also emphasizes the importance of menstrual
cycle stage in the estimation of these parameters.
The differentiation between advanced stages of the disease
(BIII/IV group) and the control group was possible from the
majority of markers used in our study because, except for
SAA at time 2 and IgG aCL at both time points, their
measurement revealed significant differences at times 1 and
2. However, the only variable that permitted a statistically
significant differentiation between early disease and control
was IgM aCL. The difficulty in obtaining markers of early
disease has only been reported in studies with CA 125 (Fedele
et al, 1988; O'Shaughnessy et al, 1993; Evers et al, 1995).
Although we demonstrated that the CA 125 II value deter-
Markers in the diagnosis of endometriosis
mined at time 1 appeared to be related to the severity of
endometriosis, from a statistical viewpoint the additional use
of CRP and IgM aCL was also relevant for 70% of the patients
with cut-off values of 33 IU/ml, and for 100% of the patients
with cut-off values of 100 IU/ml (Table III). An estimated 15
± 10% of patients were diagnosed with the disease when CRP
values were =5 3 |J.g/ml and IgM aCL values were <10 MPL,
in contrast to the 98 ± 2% of patients with CRP and IgM
aCL values >3 |ig/ml and >10 MPL respectively (Table IV).
In situations in which procedures such as laparoscopy or
laparotomy are the only definitive method of diagnosing the
disease (Evers et al, 1995; Maclavert and Shaw, 1995), the
use of serum markers becomes important to predict the presence
of the disease and aid its appropriate clinical evaluation. The
diagnostic and therapeutic indication for patients with pelvic
endometriosis can be substantiated by using determinations of
CA 125 II to predict stages III and IV of the disease, of IgM
aCL to characterize early endometriosis, and their combined
determination together with CRP rather than CA 125 II and
SAA, which are more associated with advanced stages of
the disease.
Thus the use of the invasive technique of laparoscopy can
be reduced to a minimum in cases of early endometriosis.
Conservative clinical treatment is indicated in cases of pelvic
pain, using oral contraceptives in combination with nonhormonal anti-inflammatory drugs or the direct treatment of
infertility, with laparoscopy retained as the last option when
these treatment modalities fail after 6 months or 1 year.
Measurement of these molecules may therefore provide a
valuable tool in the diagnosis and management of endometriosis.
Acknowledgements
The authors wish to thank Dalton Andrade and Julio Singer from the
Department of Statistics of Sao Paulo University, Brazil, for statistical
assistance and Dr Claudio Galperin for helping in the revision of this
manuscript.
References
Abrao, M.S., de Oliveira, R., Soares, J.B. et al. (1992) Humoral immunological
alterations in endometriosis. In Proceedings of the 3rd World Congress on
Endometriosis, Brussels, Belgium. Abstract.
Agresti, A. (1990) Categorical data analysis. John Wiley, New York, USA,
558 p.
American Fertility Society (1985) Revised classification of endometriosis.
Fertil. Steril., 43, 351-352.
Barbieri, R.L. (1990) Etiology and epidemiology of endometriosis. Am. J.
Obstet. GynecoL, 162, 565-567.
Bast, R.C., Jr, Klug, T.L., St John, E. et al. (1983) A radioimmunoassay using
a monoclonal antibody to monitor the course of epithelial ovarian cancer.
N. Engl. J. Med., 309, 883-887.
Berger, G. (1995) Epidemiology of endometriosis. In Nezhat, C.R., Berger,
G.S., Nezhat, F.R. et al. (eds), Endometriosis: Advanced Management and
Surgical Techniques. Springer-Verlag, New York, NY, USA, pp. 3-8.
Bonilla-Musoles, F. and Martinez-Molina, V. (1995) Three-dimensional
endosonography and transvaginal color Doppler in the diagnosis of
endometriosis. In Coutinho, E.M., Spinola, P. and Moura, L.H. (eds),
Progress in the Management of Endometriosis. Parthenon Publishing,
London, UK, pp. 171-173.
de Oliveira, R.M., Sipe, J.D., de Beer, F.C. et al. (1994) Rapid, sensitive
enzyme-linked immunosorbent assays (ELISA) for serum amyloid A
(apoSAA) in human plasma and tissue culture fluids (amyloid). Int. J. Exp.
Clin. Invest., 1, 23-26.
Dmowski, W.P. (1995) Immunological aspects of endometriosis. Int. J.
Gynecol. Obstet., 50, 3-10.
Evers, J.L., Dunselman, G.A. and Van Der Linden, P.J.Q. (1995) New markers
for endometriosis. In Coutinho, E.M., Spinola, P. and Moura, L.H. (eds),
Progress in the Management of Endometriosis. Parthenon Publishing,
London, UK, pp. 175-184.
Fedele, L., Vercellini, P., Arcaini, L. et al. (1988) CA-125 in serum, peritoneal
fluid, active lesions and endometrium of patients with endometriosis. Am.
J. Obstet. Gynecol., 158, 166-170.
Galle, PC. (1989) Clinical presentation and diagnosis of endometriosis. Obstet.
Gynecol. Clin. N. Am., 16, 2 9 ^ 2 .
Gleicher, N., Pratt, D. and Dudkiewicz, A. (1993) What do we really know
about autoantibody abnormalities and reproductive failure: a critical review.
Autoimmunity, 6, 115-140.
Hornstein, M.D., Harlow, B.L., Thomas, P.P. et al. (1994) Use of a new CA125 assay in the diagnosis of endometriosis. In Proceedings of the 4th
World Congress on Endometriosis, Salvador, Brazil. Abstract.
Izzo, V.M., Souza, A.Z., Mattos, A.B. etal. (1984) Importancia da laparoscopia
em ginecologia: analise de 522 casos. Ginecol. Obstet. Bras., 7, 25-35.
Jansen, R.P.S. and Russel, P. (1986) Non-pigmented endometriosis. Clinical
laparoscopic and pathologic definition. Am. J. Obstet. Gynecol., 155, 1554.
Kennedy, S.H., Nunn, B., Cederholm-Willians, S.A. et al. (1989) Cardiolipin
antibody levels in endometriosis and systemic lupus erythematosus. Fertil.
Steril, 52, 1061-1062.
Kilpatrick, D.C., Haining, R.E. and Smith, S.S.K. (1991) Are cardiolipin
antibody levels elevated in endometriosis? Fertil. Steril., 55, 436-437.
Kleinbaum, D.G., Kupper, L.L. and Morgenstern, H. (1982) Epidemiologic
Research. Belmont, Lifetime Learn Publications, London, UK.
Loose, L.D., de Oliveira, R.M., Sipe, J.D. et al. (1993) A possible systemic
component of osteoarthritis: elevated concentrations (by ELISA) of Creactive protein in serum of osteoarthritis patients and modulation by
tenidap. Arthritis Rheum., 36, S166.
Maclavert, CM. and Shaw, R.M. (1995) Pelvic pain and endometriosis. In
Shaw, R.W (ed.), Endometriosis: Current Understanding and Management.
Blackwell Science, Oxford, UK, pp. 112-146.
Neter, J. and Wasserman, W. (1974) Applied Linear Statistical Models. Richard
Irwin, Homewood, Illinois, USA.
Niloff, J.M., Knapp, R.C., Schaetzl, E. et al. (1984) CA125 antigen levels in
obstetric gynecologic patients. Obstet. Gynecol, 64, 703-707.
O'Brien, T.J., Raymond, L.M., Bannon, G.A. et al. (1991) New monoclonal
antibodies identify the glycoprotein carrying the CA 125 epitope. Am. J.
Obstet. Gynecol, 165, 1857-1864.
Odukoya, O.A., Bansal, A., Wilson, P. et al. (1996) Soluble CD23 protein in
the peritoneal fluid of patients with endometriosis. Hum. Reprod., 11,
2018-2021.
O'Shaughnessy, A., Check, J.H., Nowroozi, K. et al. (1993) CA-125 levels
measured in different phases of the menstrual cycle in screening for
endometriosis. Obstet. Gynecol, 81, 99-103.
Pittaway, D.E. and Fayez, J.A. (1987) The use of CA-125 in the diagnosis
and management of endometriosis. Fertil. Steril, 46, 790-795.
Stripling, M.C., Martin, D.C., Chatman, D.L. et al. (1988) Subtle appearance
of pelvic endometriosis. Fertil. Steril, 49, 427-430.
Weed, J.C. and Arguembourg, PC. (1980) Endometriosis: can it produce an
autoimmune response resulting in infertility? Clin. Obstet. Gynecol, 23,885.
West, C.P. (1990) Endometriosis. Br. Med. J., 301, 189-190.
Received on February 28, 1997; accepted on July 28, 1997
2527