Premium
Expression of Multiple Tumour Markers in Serum from Patients with Ovarian Carcinoma and Healthy Women
Author(s) -
Ward B. G.,
McGuckin M. A.,
Hurst T. G.,
Khoo S. K.
Publication year - 1989
Publication title -
australian and new zealand journal of obstetrics and gynaecology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.734
H-Index - 65
eISSN - 1479-828X
pISSN - 0004-8666
DOI - 10.1111/j.1479-828x.1989.tb01759.x
Subject(s) - ovarian carcinoma , ovarian cancer , antigen , stage (stratigraphy) , medicine , oncology , cancer antigen , carcinoma , gastroenterology , cancer , biology , immunology , paleontology
EDITORIAL COMMENT: Tumour markers potentially could have clinical importance if they were able to identify patients who, following treatment of invasive cancer, had residual disease and therefore required further therapy (chemotherapy). They would have a more important application if they could screen ‘normal’ populations to identify patients with subclinical disease at a stage when treatment offered greater prospects of cure. Nowhere is this concept more important than with ovarian carcinoma where the majority of patients present with late‐stage disease. Cytology of peritoneal fluid obtained by posterior colpotomy was the first screening test (other than bimanual vaginal palpation) advocated for diagnosis of early stage ovarian carcinoma — not surprisingly this test never gained popularity. Pelvic ultrasonography at the time of regular check‐up for breast and cervical carcinoma also has advocates — the increased accuracy of vaginal probe ultrasound for pelvic lesions may render this test cost‐effective in the foreseeable future, when ultrasound equipment becomes less expensive and most practitioners are trained in its use! If the incidence of ovarian carcinoma is I in 4,000 women aged 40 years or more, any screening test will need an incredible specificity (ability to reject negative cases) to be clinically useful. This paper tells us that at this time there is no useful tumour marker available for screening the general population for ovarian cancer. The CA125 assay seems to be the best available to detect antigens in the serum of patients known to have ovarian cancer. It therefore appears obvious that the CA125 assay should be tested in large populations of healthy women such as those presenting for regular check‐up for breast and uterine cancer. The surveys of 2,000 asymptomatic women referred to in this paper were not encouraging (23,24), and the authors provided the information that the CA125 assay alone costs $20.00 per sample. Theproblem of the early diagnosis of ovarian cancer remains unsolved. Twenty years ago there was a trend towards oophorectomy at the time of hysterectomy for benign diseases in women aged 40–45 years or more, the main reason being prevention of ovarian cancer. This practice was advocated by oncologists but more recently condemned by endocrinologists who emphasized that the ovarian stroma from postmenopausai ovaries synthesizes androgenic steroid hormones (testosterone, dehydroepian‐drosterone, androstenedione) important for libido and aromatization to oestrogens in peripheral fat. Both gynaecologists and patients (premenopausal and postmenopausai) are now less in favour of elective removal of normal ovaries at the time of hysterectomy because of our failure to diagnose ovarian cancer at an early stage. Summary: Serum samples from 70 patients with bulky ovarian carcinomas, 46 patients with surgically extirpated Stage I ovarian carcinomas, and 108 aged‐matched healthy control subjects were assayed for 10 tumour‐associated antigens. Levels of expression of each antigen were progressively increased in treated Stage I and bulky disease patients over healthy controls. Levels of expression in treated Stage I patients inversely reflected the interval between surgery and collection of the sample. For patients with bulky disease, determination of correlation coefficients of expression of each antigen against each other antigen showed that in 9 of 45 such relationships, the coefficients were >0.30, suggesting significant coexpression. The best correlation was found for CA125 and MSA, HMFG2 and MSA, DCA and MSA, and DCA and HMFG2. By multivariate discriminant function analysis, the combination of 2 assays (CA125 and NB/70K) was found to increase specificity of detection of ovarian carcinoma over one assay alone (CA125). Use of more than these 2 assays increased sensitivity only marginally. The attained specificity is insufficient for use as a screening assay for ovarian cancer alone, given the low prevalence in the community of this disease.