Keywords
breast cancer - breast tumor - CA 15–3 - CEA - MAM - mammaglobin
Introduction
Breast cancer is the malignant proliferation of epithelial cells that line the ducts
or lobules of the breast. Female breast cancer has now surpassed lung cancer as the
leading cause of global cancer incidence in 2020, with an estimated 2.3 million new
cases, representing 11.7% of all cancer cases. It is the fifth leading cause of cancer
mortality worldwide, with 685,000 deaths. Among women, breast cancer accounts for
1 in 4 cancer cases and for 1 in 6 cancer deaths, ranking first for incidence in the
vast majority of countries (159 of 185 countries) and for mortality in 110 countries.[1] The incidence of breast cancer deaths is on the rise and it now represents the leading
cause of cancer deaths among women in urban India.[2] Various biomarkers have been studied in breast cancer but none of them show sufficient
sensitivity and specificity to be clinically valuable.[3] CEA and CA15–3 are widely used but CEA is elevated in other cancers as well such
as colorectal cancer and pancreatic cancer.[4] Hence, there is a crucial need of finding a newer biomarker.
Human mammaglobinA (MAM) has recently been identified as a diagnostic breast cancer
marker and is almost exclusively expressed by breast tissue.[5] It is a dimeric protein belonging to the secretoglobin family and found to be associated
with lipophilin B and is secreted by breast epithelial cells. It has a transmembrane
domain and a signal peptide, which when cleaved is released into the circulation.[5] The exact function of MAM is not known but various studies prove its role in cancer
development, immune system regulation, and in the transport of aromatic molecules,
such as steroid hormones. Several studies have found that MAM is not elevated in any
other cancer.[6] As MAM is relatively a newer marker in serum and most of the studies conducted so
far have studied MAM in tissues, we intended to study the role of serum MAM as a diagnostic
marker in benign and malignant breast tumors and compare its diagnostic efficacy with
available serum markers CA15–3 and CEA.
Materials and Methods
This was a cross-sectional, case–control study comprising 27 histopathologically confirmed
breast cancer cases, 20 cases of benign breast tumors, and 30 healthy controls. Based
on the α error at 0.05 and β of 0.2, with receiver operating characteristic (ROC)
of 0.898, sample sizes in negative/positive groups as 2.15 with p-value < 0.05 from previous study.[5] The sample size calculated was 5 cases and 11 controls that justify our sample size.
The study was conducted in the Department of Biochemistry in collaboration with Departments
of Surgical Oncology and Pathology of Nizam's Institute of Medical Sciences (NIMS),
Hyderabad, from September 2018 to April 2019 after the study was approved by hospital's
institutional ethics committee (EC/NIMS/1964/2017). Informed consent was taken from
all participants. Women of age group between 18 and 75 years with breast lump visiting
the out-patient department of surgical oncology were included either in the benign
group or malignant group based on histopathological findings. Controls comprised healthy
women volunteers of similar age group. Women with a history of smoking, cancer, patients
on radiotherapy, chemotherapy, or those who underwent surgery, women below 18 years,
pregnant and lactating women were excluded from the study. Venous blood samples were
collected. Serum CA 15–3 (Roche Cobas e 411) and serum CEA (Siemens ADVIA Centaur
XP Immunoassay System) were analyzed. Serum was then aliquoted and stored at −40°C
to measure serum mammaglobin-A later using the Sandwich ELISA kit (Elabscience; Human
SCGB2A2).
Statistical analysis: Statistical analyses were performed using statistical softwaresMedCalc version 20.008
and SPSS version 25. Normality was tested using Kolmogorov–Smirnov test. Parametric
data are expressed in terms of mean and standard deviation (SD), and non-parametric
data using median and interquartile range (IQR). Pearson's correlation coefficient
was calculated for parametric data, and Spearman's coefficient was calculated for
non-parametric data. Kruskal–Wallis test was done to compare MAM and CEA among malignant,
benign, and control groups. Analysis of variance (ANOVA) was done for comparing the
means of CA15–3. The diagnostic performance of each marker in differentiating malignant
and benign breast tumor was tested using the ROC curves. For all analyses, p < 0.05 was considered statistically significant.
Results
The mean age of malignant patients was 51.2 years (31–72 years), which was significantly
higher than that of the benign group 34.4 years (19–64 years) p < 0.001. Malignant cases were either of duct cell carcinoma, invasive duct cell carcinoma,
ductal carcinoma in situ (DCIS). The most common stage at which breast cancer patients
presented was stage IIIB (45%) ([Fig. 1]).
Fig. 1 Stagewise distribution of breast cancer cases.
Serum mammaglobin levels were significantly different among the three groups, p < 0.000001 ([Table 1]). Posthoc analysis showed that the median serum MAM (ng/mL) levels were higher in
the malignant group (26.25 [23.03–27.98]) when compared with benign (11.08 [10.87–11.4],
p = 0.004) and control groups (9.2 [8.78–9.7], p < 0.000001) as seen in [Fig. 2] and also significant difference was found between benign and control groups (p = 0.0014). There was no statistically significant difference in CA15–3 levels ([Fig. 3]) among the three groups (p = 0.536). Serum CEA levels were significantly different among the three groups, p < 0.000001. The median CEA (ng/mL) levels ([Fig. 4]) were higher in the malignant group (1.85 [1.53–3.575]) when compared with the benign
(1.155 [0.915–1.705], p = 0.011) and control (0.41[0.4–0.425] p < 0.000001) groups and also significant difference was found between benign and control
groups (p < 0.00018). Serum mammaglobin has not shown any significant correlation with either
CEA (r = 0.034, p = 0.7) or CA15–3 (r = 0.08, p = 0.49).
Table 1
Serum biomarkers in healthy controls, benign and malignant breast tumor groups
|
Biomarker
|
Control group (N = 30)
|
Benign tumor (N = 20)
|
Malignant tumor (N = 27)
|
p-Value
|
|
Mammaglobin (ng/mL) Median (IQR)
|
9.2 (8.78–9.7)
|
11.08 (10.87–11.40)
|
26.25 (23.03–27.98)
|
< 0.000001*
|
|
CA 15–3 (U/mL)
Mean ± SD
|
15.06 ± 8.95
|
17.72 ± 9.00
|
17.39 ± 7.28
|
0.536
|
|
CEA (ng/mL)
Median (IQR)
|
0.41 (0.4–0.425)
|
1.155 (0.915–1.705)
|
1.85 (1.53–3.575)
|
< 0.000001*
|
Abbreviations: CA 15–3, carbohydrate antigen-15–3; CEA, carcinoembryonic antigen.
*p < 0.05 is significant.
Fig. 2 Box whisker plot depicting the distribution of mammaglobin levels in control and
breast tumor groups.
Fig. 3 Box whisker's plot depicting the distribution of CA 15-3 among three groups.
Fig. 4 Box whisker's plot depicting the distribution of CEA among the three groups. CEA,
carcinoembryonic antigen, CA 15-3, carbohydrate antigen-15-3.
The ROC analysis showed serum MAM had 81.5% sensitivity, 100% specificity, 100% positive
predictive value (PPV), and 88.9% negative predictive value (NPV) at a cut-off of
11.89 ng/mL ([Table 2]). CA15–3 showed poor diagnostic efficacy with an area of curve (AUC) of 0.555. CEA
showed 88.9% sensitivity, 82.5% specificity, 77.4%, PPV, and 91.7% NPV at a cut-off
of 1.37 ng/mL. The AUC ([Fig. 5]) was the highest for MAM (0.892), followed by CEA (0.889) and CA15–3 (0.555). The
combined ROC curve of MAM and CEA showed sensitivity of 100%, specificity of 82.5%,
and AUC of 0.913.
Table 2
Diagnostic efficacy of biomarkers
|
Serial no.
|
Parameter
|
Cut-off
|
Sensitivity (%)
|
Specificity (%)
|
PPV (%)
|
NPV (%)
|
AUC
|
|
1.
|
Mammaglobin
|
11.89 ng/mL
|
81.5
|
100
|
100
|
88.9
|
0.892
|
|
2.
|
CA 15–3
|
10.97 U/mL
|
81.5
|
35
|
45.8
|
73.7
|
0.555
|
|
3.
|
CEA
|
1.37 ng/mL
|
88.9
|
82.5
|
77.4
|
91.7
|
0.889
|
Abbreviations: CA-15–3, carbohydrate antigen-15–3; CEA, carcinoembryonic antigen.
Fig. 5 ROC curves for biomarkers mammaglobin, CA 15-3, and CEA.
Discussion
Because serum mammaglobin A is specifically produced by the breast tissue and elevated
levels are observed in breast cancer and not seen in any other cancers, this study
was done to assess the diagnostic accuracy of serum MAM in the diagnosis of benign
and malignant breast tumors and compare it with existing biomarkers.
The mean age in cancer patients was higher than in those with benign disease possibly
due to increase in the incidence of cancers with age. Due to lack of awareness or
social inhibition, most of the patients delayed their first visit to the out-patient
department after the onset of their symptoms and hence 45% of the cancer patients
belonged to stage IIIB at the time of presentation.
Serum levels of markers in patients with breast tumor
Statistically significant differences were found in serum MAM levels (ng/mL) among
malignant, benign, and control groups (26.25, 11.08, and 9.20, respectively). In accordance
with several studies,[3]
[7]
[8] including that by Bernstein et al and Zehentner et al, our study found significant
elevation in the levels of serum MAM in malignant cases when compared with benign
and healthy controls.
CA15–3 levels did not differ among the three groups. According to many studies, CA15–3
levels are low in the early stages of breast cancer but elevated levels of CA15–3
are observed in metastatic and recurrent conditions. We included only newly diagnosed
cases of breast cancer without recurrence and none of them belonged to stage IV and
therefore less chances of metastasis; thus, CA15–3 levels might not be elevated in
malignant cases when compared with controls or benign groups.
The median CEA value of malignant cases (1.85 ng/mL) was significantly higher than
that of benign cases (1.155 ng/mL) and also higher than controls (0.41 ng/mL). Mammaglobin
did not correlate significantly with the stage of cancer because of limited number
of cases ([Fig. 6]). Galvis-Jimenez too reported no correlation between MAM levels and clinical stage
of cancer.[9]
Fig. 6 Correlation of mammaglobin with stages of breast cancer.
Correlation of Mammaglobin with CA 15–3 and CEA
We also correlated serum MAM levels with existing biomarkers CA15–3 and CEA and did
not find any significant correlation with either of these markers. This might have
been because CA15–3 is more of a marker of recurrence than of primary breast cancer.
In concurrence with the findings of El Attar et al, who studied 40 Egyptian females
with primary breast cancer, we too found no correlation between MAM levels and CEA
nor with CA15–3.[10] This was explained by Antonella et al, who reported that higher levels of CA 15–3
and CEA correlated with a larger tumor burden and a more advanced disease.[11] Other tissues such as the liver and bone are capable of producing CA 15–3 CEA, whereas
MAM is breast specific. This might be the reason for poor correlation.
Diagnostic Efficacy of Three Markers
Mammaglobin showed excellent ability to detect true disease. At a cut-off of 11.89 ng/mL,
mammaglobin had a sensitivity of 81.5% and specificity of 100%. Our study correlated
well with the findings of Bernstein et al.[7] The PPV of MAM is 100% in contrast to CEA and CA 15–3. Also mammaglobin had the
highest specificity among the three markers (100%). At a cut-off of 1.37 ng/mL, CEA
had a sensitivity of 88.9% and specificity of 82.5%. When compared with CEA, CA15–3
showed a lesser sensitivity of 81.5% and poorer specificity of 35%. CA15–3 was useful
only to identify breast cancer recurrences and less often in primary cancer as confirmed
by several studies including Fejzic.[12]
ROC Curves and Area under Curve
The AUC for mammaglobin was the highest among the three markers. The AUC for CA-15–3
was the lowest. These results were consistent with the findings of Bernstein et al[7] and Galvis et al.[9] When MAM and CEA were combined, there was a slight increase in the AUC to 0.913
from their individual AUC of 0.892 (MAM) and 0.889 (CEA). The greatest advantage of
MAM over CEA is that unlike CEA, MAM is breast specific, whereas CEA is found to be
elevated in other cancers such as pancreatic, colon, and lung.[3]
[7] Mammaglobin in the malignant group is highly elevated than in the control group.
In contrast, the elevation in the benign group is not that high. Though CEA showed
significant difference among the three groups, its cut-off obtained (1.37 ng/mL) was
very low, and with the commonly used reference range cut-off of 2.5 ng/mL, we found
that the sensitivity dropped to 37% though the specificity obtained was 97%. This
makes CEA not a so useful marker. These findings highlight the valuable role of serum
MAM as a diagnostic tool in breast cancer.
Conclusion
Mammaglobin proves to be a very sensitive and specific marker of breast tumors, especially
cancer. The usefulness of MAM in diagnosis stems from its specificity in breast tissue
and its elevation in breast cancer alone, which is not so in the case of CEA. Mammaglobin
can prove as a golden alternative to mammography as a screening test for breast malignancies.
CA15–3 has lesser diagnostic accuracy in detecting primary breast cancer when compared
with MAM and CEA and must be reserved for follow-up of recurrences. Combining CEA
and MAM offers additional diagnostic efficacy in detecting cases of breast cancer.
From our observations, we can conclude that serum MAM has the potential to be used
as a diagnostic marker in breast cancer.
Limitations
A larger sample size comprising women of wider age group and tumor stages may validate
the study further.
