Patient and Tumor Characteristics

In total, the tumor punches of 557 breast cancer patients were included. Cathepsin B, L, and S expression could be successfully scored in respectively 340, 373 and 252 of these tumors—due to incidental poor tissue section quality. The clinicopathological characteristics of all patients are summarized in Table 1 (percentages are corrected for unknowns). Median age at diagnosis was 55 (IQR 47–66). Most patients (79.9%) had an invasive carcinoma no special type (NST). Tumors showed estrogen receptor (ER), progesterone receptor (PR) and human epidermal growth factor receptor 2 (HER2) overexpression in respectively 75%, 55%, and 24% of patients. Tumors were Bloom-Richardson grade 1, 2, and 3 in respectively 17%, 45%, and 38% of cases. Median tumor size was 19 mm (IQR 13–27). The majority of patients (92.6%) did not receive neoadjuvant treatment. There were no significant differences in clinicopathological characteristics between the subgroups of patients of whom the tissue was scored for a certain cathepsin (p > 0.05).

Table 1 Patient and tumor characteristicsHigh Versus Low Expression of Cathepsin B, L, And S in Breast Cancer Tissue

Figure 1a shows representative TMA punches immunohistochemically stained for cathepsin B, L, and S of tumors with respectively a high and low expression of these proteases, derived from four separate patients. All three cathepsins were expressed by breast cancer cells and cells in the stroma. Cathepsin B and L mainly showed expression by breast cancer cells. Cathepsin S was mainly expressed by cells in the stroma.

Fig. 1

(a) Representative images of cathepsin B, L, and S stained TMA punches of breast tumors. Two tumors are shown for each cathepsin: in the upper panel one with a high expression (TIS ≥ 6) and in the lower panel one with a low expression (TIS < 6). The scale bar represents 200 µm in the whole punch. The zoom area represents a 5 × magnification of the corresponding area in the whole punch. (b-d) High and low expression of respectively cathepsin B (b), L (c), and S (d) in breast tumors. In total, 340, 373, and 252 tumors were successfully analyzed for cathepsin B, L, and S expression, respectively. Shown are the percentages of tumors with certain clinicopathological characteristics that had a high (TIS ≥ 6) and low (TIS < 6) cathepsin expression. The number of tumors in each category is shown above each bar. Cathepsin B expression was significantly higher in invasive carcinomas NST compared with invasive lobular carcinomas, in ER-negative compared with ER-positive tumors, in triple-negative tumors compared with the other molecular subtypes and in higher grade tumors compared with lower grade tumors (p < 0.05). Cathepsin L expression was significantly higher in ER-negative tumors compared with ER-positive tumors, in PR-negative tumors compared with PR-positive tumors and in higher grade tumors compared with lower grade tumors (p < 0.05). Cathepsin S expression was significantly higher in invasive carcinomas NST compared with invasive lobular carcinomas, in ER-negative compared with ER-positive tumors, in PR-negative compared with PR-positive tumors, in HER2-positive compared with HER2-negative tumors, in triple-negative tumors compared with the other molecular subtypes and in higher grade tumors compared with lower grade tumors (p < 0.05). Abbreviations: CTSB, cathepsin B; CTSL, cathepsin L; CTSS, cathepsin S; ER, estrogen receptor; GR, grade; HER2, human epidermal growth factor receptor 2; ILC, invasive lobular carcinoma; LN, lymph-node; NAT, neoadjuvant therapy; NST, no special type; PR, progesterone receptor; TIS, total immunostaining score; TN, triple-negative. * = Mann–Whitney test p < 0.05. ** = Kruskal–Wallis test p < 0.05.

Cathepsin B Expression Levels

TIS of cathepsin B are summarized in Table 2. Of the total of 340 breast tumors that were successfully analyzed for cathepsin B, 95 (28%) showed a high (TIS 6–12) and 245 (72%) a low (TIS 0–5) expression. Only two tumors showed no cathepsin B expression. Cathepsin B expression was not limited to specific morphologic or molecular subtypes.

Table 2 Cathepsin B total immunostaining scores (TIS) of immunohistochemically stained breast cancer punches

Figure 1b shows the percentages of breast tumors with a high and low expression of cathepsin B, for different patient and tumor characteristics. Cathepsin B expression was significantly higher in invasive carcinomas NST compared with invasive lobular carcinomas (high in 30.5% vs. 8.6%, p < 0.05). ER-negative tumors had a significantly higher cathepsin B expression compared with ER-positive tumors (high in 40.3% vs. 25.3%, p < 0.05). Compared to the other molecular subtypes, triple-negative (TN) tumors had a significantly higher cathepsin B expression (high in 54.6% vs. 27.1%, p < 0.05). Additionally, cathepsin B expression significantly increased with advancing grade (high in 21% vs. 25.5% vs. 37.5%, p < 0.05).

There was no significant difference in cathepsin B expression between PR-positive and -negative tumors (p = 0.103) or between HER2-positive and- negative tumors (p = 0.131). Furthermore, cathepsin B expression did not differ significantly between the tumors of patients who received neoadjuvant therapy and patients who did not (p = 0.773) or between patients who had tumor-positive lymph nodes and patients who did not (p = 0.213).

Cathepsin L Expression Levels

TIS of cathepsin L are summarized in Table 3. Of the total of 373 breast tumors that were successfully analyzed for cathepsin L, 298 (80%) showed a high (TIS 6–12) and 75 (20%) a low (TIS 0–5) expression. There were no tumors without any cathepsin L expression.

Table 3 Cathepsin L total immunostaining scores (TIS) of immunohistochemically stained breast cancer punches

Figure 1c shows the percentages of breast tumors with a high and low expression of cathepsin L, for different patient and tumor characteristics. Cathepsin L expression was significantly higher in ER-negative tumors compared with ER-positive tumors (high in 89.4% vs. 80%, p < 0.05) and in PR-negative tumors compared with PR-positive tumors (high in 89.9% vs. 77.1%, p < 0.05). Additionally, cathepsin L significantly increased with advancing grade (high in 75.9% vs. 80% vs. 84.4%, p < 0.05).

There was no significant difference in cathepsin L expression between invasive carcinomas NST and invasive lobular carcinomas (p = 0.281) or between TN tumors and the other molecular subtypes (p = 0.177). Expression of cathepsin L did not differ significantly between HER2-positive and -negative tumors (p = 0.682). Furthermore, there was no significant difference in cathepsin L expression between the tumors of patients who received neoadjuvant therapy and patients who did not (p = 0.164) or between patients who had tumor-positive lymph nodes and patients who did not (p = 0.723).

Cathepsin S Expression Levels

Total immunostaining scores (TIS) of cathepsin S are summarized in Table 4. Of the total of 252 breast tumors that were successfully analyzed for cathepsin S, 46 (18.2%) showed a high (TIS 6–12) and 206 (81.8%) a low (TIS 0–5) expression. Four tumors showed no expression. Cathepsin S expression was not limited to specific morphologic or molecular subtypes.

Table 4 Cathepsin S total immunostaining scores (TIS) of immunohistochemically stained breast cancer punches

Figure 1d shows the percentages of breast tumors with a high and low expression of cathepsin S, for different patient and tumor characteristics. Cathepsin S expression was significantly higher in invasive carcinomas NST compared with invasive lobular carcinomas (high in 19.5% vs. 12.5%, p < 0.05). ER-negative tumors had a significantly higher cathepsin S expression compared with ER-positive tumors (high in 40% vs. 12.1%, p < 0.05). Similarly, PR-negative tumors had a significantly higher cathepsin S expression compared with PR-positive tumors (high in 23.6% vs. 12%, p < 0.05). Contrarily, HER2-positive tumors had a significantly higher cathepsin S expression compared with HER2-negative tumors (high in 32.6% vs. 14.3%, p < 0.05). Compared to the other molecular subtypes, TN tumors had a significantly higher cathepsin S expression (high in 45.5% vs. 15.7%, p < 0.05). Additionally, cathepsin S expression significantly increased with advancing grade (high in 21% vs. 25.5% vs. 37.5%, p < 0.05).

There was no significant difference in cathepsin S expression between the tumors of patients who received neoadjuvant therapy and patients who did not (p = 0.467) or between patients who had tumor-positive lymph nodes and patients who did not (p = 0.929).

Association of Cathepsin B, L, and S Expression Within the Same Breast Tumor

Due to incidental poor tissue section quality, not all tumors could be analyzed for the expression of each cathepsin. Of all 557 included tumors, 173 were successfully stained and analyzed for both cathepsin B and L, 149 were successfully stained and analyzed for both cathepsin B and S, 132 were successfully stained and analyzed for both cathepsin L and S, and 46 were successfully stained and analyzed for all three cathepsins.

All tumors showed cathepsin B, L, and/or S expression to some extent (TIS > 0). In 41 (89%) of the 46 tumors scored for all three cathepsins, the expression level of one or more of these proteases was scored as high (TIS > 6). A weak but significant correlation was found between the expression levels of cathepsin B and L (ρ = 0.18, p < 0.05), cathepsin B and S (ρ = 0.33, p < 0.01), and cathepsin L and S (ρ = 0.31, p < 0.01).