INTRODUCTION
Breast cancer is a life-threatening malignancy which is most common among
women and the second leading cause of cancer death in women today (Elzagheid,
et al., 2006). The earlier this cancer is detected, the earlier
the opportunity for treatment and hence the lower the risk of death. Currently,
women suspected of having breast cancer are most likely to have a biopsy
taken for histological investigation, since only a histological examination
of biopsy can confirm or rule out cancer. But a large number of people
who undergo biopsy do not have cancer, yet they go through the anxiety
and trauma of the procedure and its complications afterwards.
Numerous serum tumor markers have been described for breast cancer, including
members of the MUC1 family of mucin glycoproteins (e.g., CA 15-3, BR 27.29
and MCA, CA 549), Carcino Embryonic Antigen (CEA), oncoproteins (e.g.
HER-2/c-erbB-2) and cytokines (e.g., tissue polypeptide antigen and tissue
polypeptide-specific antigen) (Tondini et al., 1989).
It is generally agreed that tumor markers in breast cancer patients are
not a tool for primary diagnosis because of their low sensitivity and
specificity (Tondini et al., 1989).
The carbohydrate antigen CA 15-3 however, has been proven to be the most
sensitive and specific tumor marker for breast cancer. But there have
been differing views on its sensitivity and specificity to recommend it
for screening, diagnosis, staging, or surveillance after primary treatment
for breast cancer [American Society of Clinical Oncology (ASCO) Tumor
Markers Expert Panel, 2001].
Nevertheless, it is used to detect distant recurrence after primary treatment
in breast cancer patients, where it often increases before clinical symptoms
become evident (Mohammad Nejad et al., 2006). The aim of this study
was to determine the specificity and sensitivity of serum CA 15-3 for
the detection of breast cancer in Kumasi, Ghana by comparing the levels
of serum CA 15-3 in breast cancer patients before and after treatment
and also in suspected breast cancer patients with the histopathological
findings.
The objective was to determine the levels of serum CA 15-3 to enable
early and non-invasive detection of breast cancer in the preclinical state
and to monitor treatment of breast cancer patients.
A total number of 110 female patients who visited the Peace and Love
Clinic at Oduom, in Kumasi from November, 2007 to February, 2008 were
involved in the study. These consisted of 35 known previously confirmed
(by histological examination) breast cancer patients and 75 suspected
(unconfirmed) breast disease patients. Twenty (20) apparently healthy
females were used as controls. Patients examination, specimen collection
(both biopsy and blood specimens) were performed with patients consent
and in accordance with the Helsinki Declaration. Each patient was given
a questionnaire to complete.
About 3 mL of venous blood sample was taken from each of the 110 patients
as well as the 20 healthy controls. All samples were processed to obtain
sera. For the 35 confirmed breast cancer patients, samples were collected
before and three months after treatment (surgery, chemotherapy and radiotherapy).
Serum samples were kept at 20°C and were brought to room temperature
before assay.
Quantitative determination of serum CA 15-3 concentrations were made
using BioCheck CA 15-3 Enzyme Immunoassay. Values greater than 35 U mL-1
were considered abnormal.
Tissue biopsies were surgically removed by a qualified surgeon from the
75 unconfirmed breast disease patients and sent to the histopathology
laboratory for processing. The processed tissues were stained with Haematoxylin
and Eosin (HE). Microscopic examination of each stained sample was made
by a pathologist. The histopathological findings were documented for each
patient.
Statistical analyses were performed using SPSS 15.0 statistical software
and GraphPad Prism 5 for Windows. The Paired-Sample t-test and Independent-Sample
t-test were used to compare the serum CA 15-3 concentrations, the age,
histopathological results and the menopausal status. Results were considered
significantly different at p<0.05. The Receiver Operating Curve (ROC)
analysis was used to evaluate the reciprocal relationship between sensitivity
and specificity of the serum CA 15-3 using the GraphPad Prism 5. A cut-off
value of 35 U mL-1 was used. Results were obtained at 95% confidence
interval.
The mean age of the 110 subjects used in the study was 46.2±12.0
years with minimum and maximum ages being 21 and 80 years, respectively.
Out of the 110 subjects, 59(53.6%) were of post-menopausal status and
51(46.4%) were of pre-menopausal status. Also, 35(31.8%) of them were
known breast cancer patients and the other 75(68.2%) were suspected cases.
Out of the 35 known breast cancer patients 11(31.4%) of them had normal
serum CA 15-3 levels and the other 24(68.6%) had higher than normal serum
CA 15-3 levels. After treatment, 32(91.4%) out of the 35 known cases had
normal serum CA 15-3 levels and the remaining 3(8.6%) still had higher
serum CA 15-3 levels.
Out of the 75 suspected cases, 38(50.7%) had normal serum CA 15-3 levels
and the remaining 37(49.3%) had values above normal. The histopathological
results of the biopsies demonstrated that 29(38.7%) out of the 75 suspected
cases had benign breast disease and 46(61.3%) had breast (malignant) cancer.
Of the 29 benign cases, 27(93.1%) had normal serum CA 15-3 levels, with
the remaining 2(6.9%) subjects having values above normal. With the 46
malignant cases, 11(23.9%) had normal serum CA 15-3 levels and 35(76.1%)
of them had values higher than normal.
The mean values of serum CA-153 for the various categories of patients
are shown in Table 1. There was a significant difference
between the serum CA 15-3 concentrations of patients before and after
treatment (p<0.05) (Table 2). There was also a strong
positive correlation between the level of serum CA 15-3 and the histopathology
results of the biopsies (r = 0.518).
Table 1: |
Mean levels of serum CA 15-3 for the various breast
disease categories (U mL-1)
|
|
N = Total No., SD = Standard Deviation, SEM = Standard
Error Mean, Min. = Minimum, Max. = Maximum |
Table 2: |
Paired sample correlation of serum CA 15-3 for the various
categories of patients |
|
N = Total No., Sig. = Significant value |
Table 3: |
Paired sample test CA 15-3 of patients with malignant
and CA 15-3 of patients with benign histopathology |
|
|
Fig. 1: |
ROC curve of serum CA 15-3 of controls and breast cancer
patients |
There was no correlation between the serum CA 15-3 concentrations and
the menopausal status of the patients (r = 0.185). Similarly, there was
no correlation between the age of patients and the serum CA 15-3 concentration
of patients (r = 0.209).
There was a significant difference between the serum CA 15-3 of patients
with malignant histopathology and the serum CA 15-3 of patients with benign
histopathology (p = 0.000), by Paired-Sample t-test (Table
3).
The ROC analysis of the serum CA 15-3 concentrations of the control subjects
and those of the patients with breast cancer (malignant histopathology)
showed the serum CA 15-3 to have a sensitivity and specificity of 76.1
and 100%, respectively at a cut-off of >35 U mL-1 (Fig.
1).
One area under intense research is a search for an ideal tumour marker
for the detection of breast cancer when the tumour is still small. At
the moment, only a histological examination of a biopsy can confirm or
rule out cancer. This study has shown a strong positive correlation between
the serum CA 15-3 concentrations of patients and their histopathology
results (r = 0.518). There was also a significant difference between the
serum CA 15-3 of patients with breast cancer and the serum CA 15-3 concentrations
of the patients with benign breast disease (p<0.05). A study conducted
by Lee et al. (2004), however, found no significant difference
between them.
There was a significant difference between the serum CA 15-3 concentrations
before and after treatment (p<0.05). In monitoring a patient longitudinally,
small increases or decreases in tumor marker concentrations can be indicative
of early recurrence of the disease or response to therapy, respectively
(Cheli et al., 1998). It can therefore be said from this study
that, serum CA 15-3 can be a useful marker in the monitoring of treatment
for breast cancer.
However, there was no significant correlation between the age and the
concentrations of serum CA 15-3 of the patients (r = 0.209). This observation
is consistent with the study done by Keyhani et al. (2005). Nevertheless,
a study by Lumachi et al. (2000), found a correlation between the
age and the CA 15-3 levels in the patients.
Also there was no significant difference between the serum CA 15-3 concentrations
of the patients at the post-menopausal stage and that at the pre-menopausal
stage (p>0.05). This observation is consistent with the study done
by Vizcarra et al. (1996). However, the study conducted by Duffy
(1999), found a significant difference between the concentrations of serum
CA 15-3 in post-menopausal patients and that of pre-menopausal patients.
The sensitivity and specificity of serum CA 15-3 in detecting breast
cancer using a cut-off value of 35 Um L-1 was 76.1 and 100%,
respectively. However, from the research of Keyhani (2005) sensitivity
and specificity of serum CA 15-3 were 14 and 92.3%, respectively. This
disparity may have occurred due to the different assay procedure and a
cut-off value of 30 U mL-1. Since sensitivity and specificity
are both high in this study, serum CA 15-3 can be considered to be of
help in early detection of breast cancer.
In conclusion, serum CA 15-3 can be used in preliminary screening for
breast cancer and can also be used for the monitoring of treatment and
the detection of distant recurrences after primary treatment.
In screening for breast cancer, patients with abnormal concentrations
would be considered as having the cancer and those with concentrations
within normal range would have a histological examination of their biopsy
tissues for confirmation. Also, serial sampling for determination of concentrations
of serum CA 15-3 would be done, when monitoring for efficacy of treatment,
as serum CA 15-3 has a long half-life.
However, further prospective studies, using greater numbers of patients
are required before serum CA 15-3 can be recommended for routine clinical
use.
ACKNOWLEDGMENT
We wish to express our gratitude to Samuel Kwakye Yeboah who helped in
the typing of this article.