Currently used tumor biomarkers related to BC diagnosis and management are CA15-3, CA27.29, CEA, and CA125 [21–23]. They can be used to monitor the progression of cancer, evaluate the outcome of specific treatment, or monitor a recurrence. However, tumor biomarker tests have their own limitations; the level change of the biomarker is not always caused by cancer, and healthy subjects quite often also show increased levels of certain biomarkers. Therefore, tumor biomarker tests cannot be used alone to evaluate or manage BC. Despite the clinical benefits of using tumor biomarker tests to assist in the correct reading or compensate for technical limits of imaging examinations, it is difficult to find ones with sufficiently high sensitivity and specificity.
In mammalian cells, Trx1 is involved in the regulation of reactive oxygen species (ROS) levels [28, 29]. As it plays a role in the regulation of cellular redox homeostasis, Trx1 has multiple functions in the cell. Therefore, Trx1 is an important entity that is potentially related to the onset of many diseases, including cancer, inflammation diseases, heart failure, and so on. Trx1 has been known to play an important role in regulating cancer cell growth by modulating the DNA binding activity of transcription factors [29–33]. It has been reported that the blood level of Trx1 was specifically higher in breast cancer compared with a few other cancer types [24]. Even though Trx1 is likely expressed in a few different types of cancer as well as in normal healthy women, the largest difference in the Trx1 level between sera from normal and cancer subjects was shown in BC. Therefore, if this Trx1 level difference could be distinguished from the difference engendered by normal healthy women and other types of cancers, it would be possible to detect BC from the blood. This was the main basis of BC-specific detection of Trx1 in the present study.
When the cut-off value was set at 14.13 Unit/ml of Trx1, the sensitivity and specificity for BC detection was 94.3% and 94.7%, respectively with an AUC of 0.985. These values were relatively high compared to previously reported values for other protein cancer biomarker tests, which ranged from 60–90% [21–23]. The average value of Trx1 level in normal sera was 5.45±4.16 U/ml and that of sera from BC patients was 21.96±6.79 U/ml, which was a large enough difference to identify BC subjects from normal healthy ones. The Trx1 level of patients with other types of cancer was close to that of normal healthy subjects, indicating that the blood Trx1 level could discern BC from other types of cancer. Even though the sample numbers of certain other types of cancer were insufficient, it still provides important clues about what could be expected from those other types of cancer. It should be acknowledged that there is a pile of reports showing increased expression of Trx1 in various types of cancer [34–39]. However, most of the studies were carried out in cancer cells or tissue not in blood, and focused on the gene expression level of Trx1. Since the physiological environment of blood is quite different from that of cells or tissue, it is necessary to scrutinize the level of Trx1 protein in blood from large numbers of patients with different types of cancer. The effect of many chronic inflammation diseases was also tested, and it did not show any meaningful effect on the ability of Trx1 level to identify BC patients (data not shown).
Breast cancer patients were identified by the level of Trx1 in sera, and this was not influenced by age. In particular, BC patients in their late 40s to early 50s, the age group with the highest BC incidence in Korea, were differentiated from normal subjects. Breast cancer patients in other age groups were also equally well identified by Trx1 level. It has also been reported that there is no influence of the menstrual status of Caucasian women on Trx1 level [24]. It was interesting that the level of Trx1 did not significantly change with subject age. Since it has been commonly accepted that the incidence rate of BC gets higher in older age [2, 3], it could be expected that Trx1 level would follow suit. However, no significant difference was observed.
Breast cancer has complicated classifications and shows its own pathological characteristics for diagnosis and treatment as well [16, 17]. IDC, which comprised the largest number of BC patients in the study and in the Korea breast cancer incidence database, had an average Trx1 level of 22.00±7.04 U/ml. ILC and MC also showed similar values, indicating that blood Trx1 level can detect BC regardless of BC type. The number of cases of DCIS, IMC and ITC were low, since the incidence rate of each was low and, thus, it was difficult to collect blood from these types of BC patients. All DCIS blood showed a high value of Trx1, which implies the potential of the Trx1 level for the very early detection of BC. It will be interesting to conduct a study with a large numbers of DCIS blood samples to check whether this is the case. If the level of Trx1 can detect BC even in large number of DCIS cases, it will be useful for the early detection of BC. Despite different pathological characteristics of different types of BC, there was no significant difference between Trx1 levels in any type. This strongly indicates that the blood level of Trx1 is a good means to detect BC.
As it was likely that the blood level of Trx1 was suitable to detect BC, it was interesting to check whether it could differentiate the stage of BC. The level of Trx1 in blood was not significantly changed by the stage of BC. It showed higher values than the cut-off value no matter what the stage of BC was, and thus, it indicated that Trx1 level was likely to detect BC regardless of BC stage. It was promising to see high accuracy to identify patients from stage 1 as well as other stages. This result means that the blood level of Trx1 could be an alternative modality for the early detection of BC.
The sensitivity (68.9%) and specificity (100.0%) of mammography of BC in this study is typical when compared with other recent reports on mammography [11, 13]. However, the fact that more than one third of biopsy confirmed BC patients in this study were judged as an inconclusive (BI-RADS category 0) is concerning. The BI-RADS category 0 means that the final assessment of the mammogram is likely to be held off until additional tests and images are available. It may cost money, time, and anxiety. It is interesting that the Trx1 test was generally superior to mammography in this study, and that the combined test with mammography and Trx1 showed the highest accuracy to detect BC. Since biomarker tests cannot be used alone to diagnose a certain cancer or disease, and mammography has not achieved desirable performance for BC screening, the combined test will yield complementary effects that mitigate the weakness of each test and provide the highest accuracy for BC detection.
Although there is no doubt of the need to carry on a larger population study with different types of cancer, it is likely that the level of Trx1 can detect BC specifically. In addition, Trx1 level was not affected by different characteristics of BC. This indicates that the blood level of Trx1 has potential as a biomarker for BC. Although there have been many reports regarding the role or mode of action of Trx1 in cancer, it has not been completely elucidated [40–44]. When cancer cells grow, the microenvironment of cancer cells is in a hypoxic state that favors ROS generation, resulting in higher oxidative stress. Cancer cells are inclined to protect themselves from oxidative damage via maintaining their redox status to survive and metastasize to distant organs. In this condition, Trx1 modulates redox signaling pathways via thiol-disulfide exchange with redox-responsive proteins, such as the transcription factors Ref-1 and NF-κB, MAP3K5/ASK1, and Trx1 interacting protein. This kind of signaling causes modulation of cell kinetics such as activation of proliferation, inhibition of apoptosis, and facilitation of metastasis of cancer cells. However, most previous studies have been done with cancer cell lines or cancer tissue that had an environment very different to the blood system, especially in terms of redox status. It has been reported that the mRNA level of the Trx1 gene is correlated with the proceeding of BC stages, whereas the protein level of Trx1 in blood is likely kept constant [24]. It can be assumed that Trx1 is also likely involved in other cell kinetics, such as the initiation or switch mechanism of BC. It requires a well-designed elaborate study to prove this kind of speculation.
It has been long accepted that mammography contributes to the early detection of BC, thus lowering its mortality. Nevertheless, it has some well-known limitations, such as lack of mobility, radiation exposure, uncomfortable personal experience, and cost in certain countries. In addition, satisfactory sensitivity and specificity of mammography, especially for women with dense breasts, has not been reached. Therefore, it would be greatly beneficial to have a simple, economic, and complementary means to detect BC with high accuracy in relatively early stages. When the Trx1 level of patients was analyzed along with readings of mammograms from the same patients, the sensitivity and specificity of coupled tests went up to 99.25% and 100.00%, respectively. Interestingly, this benefit was more obvious in the case of dense breasts (manuscript in preparation). Therefore, it seems that the level of Trx1 in blood could be a promising modality to detect BC as a complement to current diagnostic methods.