Changes in Cytokeratin 18 during Neoadjuvant Chemotherapy of Breast Cancer: A Prospective Study

Background & Objective: Prediction of response to neoadjuvant treatment is an important part of treatment of patients with breast cancer. This study aimed to assess changes in serum levels of Cytokeratin 18 during neoadjuvant chemotherapy in patients with locally advanced breast cancer and its association with neoadjuvant treatments. Methods: This research was performed on newly diagnosed breast cancer patients referred to Omid Radiotherapy Center and radiotherapy and oncology departments of Emam Reza and Ghaem hospitals, in Mashhad, Iran. Serum levels of M30 and M65 fragments of Cytokeratin 18 were measured before and 24 hours after the first course of neoadjuvant chemotherapy. Changes in serum levels of Cytokeratin 18 and its fragments and their correlation with pathologic response were analyzed. Results: Pre- and post-chemotherapy levels of M30 were respectively 223.9±18.94 and 250.7±23.92 U/L (P=0.24). For M65, these levels were respectively 301.5±313.9 and 330.2±352.2 U/L (P=0.1). Changes in M30 level during chemotherapy in patients with and without pathologic complete response were -20±92.69 and 43.1±106.5, respectively (P=0.1). For M65, these changes were respectively -247±55 and 76±240 (P=0.1). Baseline levels of M30 and M65 had no relation with menopausal status, tumor grade, hormone receptor status, Ki67 expression, molecular subtype, and stage. Conclusion: Our findings showed statistically insignificant changes in the level of Caspase-cleaved- (M30) and uncleaved- (M65) cytokeratin 18 fragments (apoptotic and necrotic indicators, respectively) during neoadjuvant chemotherapy in patients with breast cancer. There was no notable relationship between tumor-related factors and either baseline levels or serum changes of CK18 fragments. Also, there was no correlation between M30/M65 level and pathologic response to neoadjuvant chemotherapy.


Introduction
Breast cancer is one of the most commonly diagnosed neoplasms in women (1,2). In recent decades, remarkable progress in systemic chemotherapy, hormone therapy, and targeted therapy has significantly reduced the mortality of stage I, II, and III breast cancer patients. Given the benefits observed in adjuvant treatments, recent years has seen growing interest in the use of neoadjuvant treatments for earlier inhibition of micrometastasis in order to improve the treatment outcome (3). Today, neoadjuvant chemotherapy treatment is widely accepted as a standard treatment for locally advanced and inflammatory breast cancers. Research has shown that patients whose primary tumor exhibits a pathologic complete response (pCR) to neoadjuvant treatment generally have better treatment outcomes and survival rates (4). Assessment of response to neoadjuvant treatment is of paramount importance for preventing excess toxicity in patients who do not benefit from chemotherapy. For most patients, this assessment only involves clinical Vol. 15 No.2 Spring 2020 IRANIAN JOURNAL OF PATHOLOGY examination or imaging to track changes in the primary tumor and adjacent lymph nodes (5). Currently, there is no standard method for assessment of the response to neoadjuvant chemotherapy in breast cancer patients; however, several studies have shown that mammography, ultrasound, and MRI examinations may have a role in the prediction of pCR (5)(6)(7)(8). However, there is still no standard protocol for a definitive diagnosis of the response of primary tumors to neoadjuvant chemotherapy. Cytokeratins are among the most important components of the epithelial cell structure and have unique expressions in different epithelial cells. Two important cytokeratins of the intermediate filament family are cytokeratin 18 and cytokeratin 8, which belong to the group of type I and type II cytokeratins respectively. Together, these cytokeratins play a notable role in important cellular processes such as proliferation and cellular mitosis. During apoptosis and necrosis processes, caspase-cleaved-(M30) and uncleaved-(M65) cytokeratin 18 fragments appear at an elevated level in blood serum. Therefore, serum levels of these fragments can perhaps be used as a marker for diagnosis and to predict prognosis or response to treatment (9)(10)(11)(12)(13). In a study by Stoetzer et al., the possibility of using the level of cytokeratin 8/18 fragments (M30 level) as a specific marker for apoptosis in healthy subjects (146 U/L), patients with benign breast tumors (131 U/L), patients with local breast cancer (165 U/L), and patients with metastatic breast cancer (198 U/L) was studied. Ultimately, this study reported that only the M30 level of metastatic patients was significantly higher than that of healthy individuals (12). In a study carried out by Demiray et al., where the monoclonal level of M30 antibody was measured as the representative of cytokeratin 8/18 fragment, the M30 levels of breast cancer patients 24 hours (100/5 U/L) and 48 hours (86/5 U/L) after neoadjuvant chemotherapy were significantly higher than the baseline level (64/9 U/L) (9). Also, the change in M30 level 24 hours after chemotherapy was significantly higher in responders (increase in median M30 level from 63 U/L to 79 U/L with P=0.001) than in non-responders (increase in median M30 level from 80 U/L to 105 U/L with P= 0.05). However, a study by Fehm et al. reported that although the pathologic response of breast cancer patients could be predicted by apoptotic DTC (in the sense that none of the patients with progressing disease had rising M30 level), some of patients with a pathologic complete response (and elevated marker level) could still have non-apoptotic cells and therefore benefit from secondary adjuvant chemotherapy (10). A study by Tas et al. reported that although the serum M30 level was higher in metastatic patients, it was not significantly related to the prognosis and survival. This study also found no correlation between the serum M30 level and the response to chemotherapy (13). The present study aimed to investigate the relationship between baseline level and serum changes of cytokeratin 18 fragments M30 and M65 before and after neoadjuvant chemotherapy with pathologic response in breast cancer treatment.

Results
A total of 35 patients entered the study. The mean age of the patients was 46.17±10 years. Most of the patients were pre-menopausal (22 cases, 62.9%). The basic information of the patients is provided in Table  1. No correlation was observed between tumor characteristics and M30 and M65 levels and their changes during chemotherapy (Figures 1 and Fig2).
The pre-and post-treatment serum levels of M30 were respectively 223.9±18.94 and 250.7±23.92 U/L (P=0.24). For M65, the serum levels before and after treatment were respectively 313.9±50.96 and 356.2±55.82 U/L (P=0.1). The pre-and post-treatment M30/M65 ratio was respectively -1.07±0.14 and -0.95±0.10 (P=0.4). No relationship was found between changes in these tumor markers and survival statistics ( Figure 3). The mean change in the M30 level for the patients with pathologic complete response to chemotherapy (n=9) was -20±92.69, and for the patients without complete response (n=26), it was 70.06±240 (P=0.2) (Figure 3).

Discussion
Breast cancer is the most commonly occurring cancer in women. The biomarkers traditionally used in predicting the prognosis of breast cancer and its response to treatment include the expression of estrogen and progesterone receptors and the rate of HER2 gene amplification. Given that each molecular subtype of breast cancer itself consists of several heterogeneous sub-branches, there is a growing need to discover new biomarkers for predicting the most effective treatment approaches. Since personalized medicine and neoadjuvant chemotherapy have become increasingly popular treatment modalities in different malignancies, it is important to find a marker for early identification of those patients who will benefit from this approach (14)(15)(16). Response to neoadjuvant treatments may be predicted by cell death markers such as cytokeratins (17)(18)(19). This study investigated the relationship of serum CK18 levels before and after neoadjuvant chemotherapy with pathologic response to breast cancer. The results showed that the patients with tumors of higher grades had higher pre-treatment M30 and M65 levels. The hormone-receptor-negative patients had higher serum M30 levels and the hormone-receptorpositive patients had higher serum M65 levels. The patients with a higher level of proliferation and Ki67 expression were found to have a higher level of M30, but in contrast, when Ki67 expression was lower, the M65 level was higher. However, none of these differences was statistically significant. Table 2 shows the general information of some of the studies performed on different aspects of cytokeratin 18 in breast cancer patients. Most of the studies in this area have measured the changes in M30 level of breast cancer patients and the studies on the changes in M65 or M30/M65 ratio are less frequent.
In the present study, the results showed increased serum M30 and M65 levels after neoadjuvant chemotherapy, but the observed changes were not statistically significant. This is consistent with the majority of existing reports in this area, although there has been some discrepancy in the exact change in serum levels of this marker following chemotherapy treatment. For example, it has been reported that for breast cancer patients, docetaxel-containing treatments are associated with a significant increase in the serum level of M30, but treatment with cyclophosphamide/epirubicin/5FU results in heterogeneous responses are often accompanied with increased M65 levels (20). Given that the M30 subunit of cytokeratin 18 is released after apoptosis and M65 is released after necrosis, the serum levels of these markers can be expected to increase following chemotherapy. Similar   (9,20). Demiray's study, which was conducted on 42 breast cancer patients undergoing neoadjuvant chemotherapy with anthracycline, found a significant relationship between changes in M30 level and the overall tumor response to neoadjuvant chemotherapy, in the sense that patients whose tumors responded to chemotherapy had significantly increased M30 levels. However, it should be noted that the patients whose tumors did not respond to chemotherapy also showed an insignificant increase in this variable (9). The study performed by Olofsson (2007), which used cytokeratin 18 to monitor the cell death induced by each course of chemotherapy with docetaxel or ECF in 61 breast cancer patients, reported a significant relationship between increased serum M30 level and the pathologic response to these chemotherapies (20). Concerning the relationship between serum M30 and M65 levels and clinicopathologic factors, the results of this study showed higher M30 and M65 levels in patients with higher grade tumors, triple negativity, and more node involvement. Also, higher M30 levels were observed in patients with hormone-receptor-negative subtype and higher proliferation, and higher M65 levels were seen in patients with hormone-receptorpositive subtype and lower proliferation. However, the observed differences were not statistically significant. Previous studies have reported contradictory results in this regard. In the research performed by Tas Krawczyk et al., where the incidence of M30 in breast tumor tissue was investigated by immunohistochemistry and the results showed higher incidence of this cytokeratin in hormone-positive and HER2-positive subtypes and cancers with lower grade tumors (23). The results of the present study showed no significant relationship between changes in M30, M65 and M30/M65 ratio during chemotherapy and survival parameters. Other studies have also reported that the serum levels of M30 and M65 and their changes during neoadjuvant chemotherapy have no prognostic value for predicting survival in patients with breast cancer (13). The most important limitation of this study was the small size and high heterogeneity of the sample studied. This heterogeneity was present not only in tumor biology but also in patients' therapeutic regimens (24). Future studies are therefore suggested to use a larger sample of breast cancer patients with higher homogeneity in terms of molecular criteria. It is also recommended to investigate the impact of other interventions such as surgery and bulk tumor removal on the serum level of the studied markers. A longer assessment of this serum marker during neoadjuvant chemotherapy (repeating measurements 48 and 72 hours after treatment and after each course of chemotherapy) may also provide a better insight into the physiology of M30 and M65 (25).

Conclusion
This study found statistically insignificant changes in serum M30 and M65 levels (cleaved and uncleaved fragments of cytokeratin 18) following neoadjuvant chemotherapy in breast cancer patients. No relationship was observed between the baseline level of this marker before treatment and tumor characteristics, which indicates the lack of prognostic value. There was also no relationship between changes in serum M30 and M65 levels and pathologic response of breast tumors to chemotherapy. According to these findings, M30 and M65 cannot serve as accurate markers for predicting breast cancer response to neoadjuvant chemotherapy.