Introduction: Tumor cells use adhesion molecules like CD15 or sialylCD15 (sCD15) for metastatic spreading. We analyzed the expression of CD15 and sCD15 in clear cell renal cell carcinoma (ccRCC) regarding prognosis. Methods: A tissue microarray containing tissue specimens of 763 patients with ccRCC was immunohistochemically stained for CD15 and sCD15, their expression quantified using digital image analysis, and the impact on patients’ survival analyzed. The cell lines 769p and 786o were stimulated with CD15 or control antibody in vitro and the effects on pathways activating AP-1 and tumor cell migration were examined. Results: ccRCC showed a broad range of CD15 and sCD15 expression. A high CD15 expression was significantly associated with favorable outcome (p < 0.01) and low-grade tumor differentiation (p < 0.001), whereas sCD15 had no significant prognostic value. Tumors with synchronous distant metastasis had a significantly lower CD15 expression compared to tumors without any (p < 0.001) or with metachronous metastasis (p < 0.01). Tumor cell migration was significantly reduced after CD15 stimulation in vitro, but there were no major effects on the activating pathways of AP-1. Conclusion: CD15, but not sCD15, qualifies as a biomarker for risk stratification and as an interesting novel target in ccRCC. Moreover, the data indicate a contribution of CD15 to metachronous metastasis. Further research is warranted to decipher the intracellular pathways of CD15 signaling in ccRCC in order to characterize the CD15 effects on ccRCC more precisely.

Distant metastasis is a common event associated with malignancies. Metastases are either already present at the time of diagnosis (synchronous metastasis) or can occur at distant sites during clinical course (metachronous metastasis). Approximately 33% of patients with renal cell carcinoma (RCC) present with synchronous metastasis, and patients with initially localized tumors develop metachronous metastasis in approximately 25% [1]. Metastases have an enormous impact on patients’ survival: 5-years survival rate in nonmetastasized patients with RCC is 85%, whereas it drops to 10% in patients with metastases [2].

To form distant metastases, tumor cells must gain access to the blood system, adhere to the endothelial layer, and finally extravasate through the vessel wall. Here, the interaction of cell adhesion molecules like e-selectin and its ligand sialylCD15 (sCD15) is important as they mediate rolling of circulating tumor cells on the vessel wall [3‒6].

CD15 is an oligosaccharide moiety on glycolyzed cell surface molecules, closely related to sCD15. Both molecules share repeating N-acetyllactosamine structures as a common precursor. For synthesis of CD15, these n-acetyllactosamine structures are fucosylated by α-(1,3)-fucosyltransferase. Synthesis of sCD15 requires an intermediate step with the addition of a sialyl group before fucosylation [7, 8]. Alternatively, CD15 can be synthesized from sCD15 through sialydase-mediated desialysation [9]. Next to myeloid cells, CD15 is widely expressed in normal tissues, e.g., colonic or oral mucosa as well as breast parenchyma [10]. Functionally, CD15 is discussed in the context of neutrophil adhesion to the vascular endothelium [11, 12] and was shown to be involved in selectin-binding of Hodgkin cells [13]. In RCC, CD15 is suggested to be helpful in distinguishing several RCC subtypes [14]. Lack of CD15 in localized clear cell renal cell carcinoma (ccRCC) has been found to significantly correlate with shorter overall survival (OS) [15]. In contrast, expression of sCD15 in RCC is strongly associated with advanced tumor stadium and disease recurrence [16, 17], increased invasiveness, and unfavorable prognosis [18].

So far, most studies have focused on sCD15 in RCC. However, the prognostic values of CD15 and sCD15 in ccRCC have not been compared, especially including advanced tumor stages, and the contribution of CD15 to metastatic spread has not yet been investigated.

Patients

A tissue microarray (TMA) with samples from 763 patients with primary ccRCC, treated at the Department of Urology at the University of Heidelberg between 1987 and 2005, was used (Table 1). To account for tumor heterogeneity, two cores from different tumor areas per case were included in the TMA. Clinical follow-up was available for all cases. Survival was calculated from the date of diagnosis to three different events: disease-specific survival (DSS, event = tumor-related death), OS (event = death by any cause), and progression-free survival (PFS, event = recurrence, metastasis, or tumor-related death). Patients who did not experience the investigated events were censored. Further details have been described previously [19, 20]. Distant metastasis at the time of diagnosis was termed synchronous metastasis and distant metastasis during clinical course was termed metachronous metastasis [21]. The tumors were graded according to the three-tiered nuclear grading system [22].

Table 1.

Summary of clinicopathological features

CharacteristicN = 763*
Grade (G), n (%) 
 G1 215 (28.2) 
 G2 411 (53.9) 
 G3 133 (17.4) 
 Missing 4 (0.5) 
Tumor extent (T), n (%) 
 T1 417 (54.7) 
 T2 61 (7.9) 
 T3 260 (34.1) 
 T4 25 (3.3) 
Local lymph node metastasis (N), n (%) 
 N0 718 (94.1) 
 N1 45 (5.9) 
Distant metastasis (M), n (%) 
 M0 642 (84.1) 
 M1 121 (15.9) 
Sex 
 Female 297 (38.9) 
 Male 466 (61.1) 
Age at surgery, years, n (%) 
 <65 436 (57.1) 
 ≥65 327 (42.9) 
Karnofsky performance index, n (%) 
 >80% 463 (60.7) 
 <80% 300 (39.3) 
CharacteristicN = 763*
Grade (G), n (%) 
 G1 215 (28.2) 
 G2 411 (53.9) 
 G3 133 (17.4) 
 Missing 4 (0.5) 
Tumor extent (T), n (%) 
 T1 417 (54.7) 
 T2 61 (7.9) 
 T3 260 (34.1) 
 T4 25 (3.3) 
Local lymph node metastasis (N), n (%) 
 N0 718 (94.1) 
 N1 45 (5.9) 
Distant metastasis (M), n (%) 
 M0 642 (84.1) 
 M1 121 (15.9) 
Sex 
 Female 297 (38.9) 
 Male 466 (61.1) 
Age at surgery, years, n (%) 
 <65 436 (57.1) 
 ≥65 327 (42.9) 
Karnofsky performance index, n (%) 
 >80% 463 (60.7) 
 <80% 300 (39.3) 

*n (%).

Immunohistochemistry

Antibodies directed against CD15 (ready-to-use, Carb-3, IR062, Dako, Glostrup, Denmark) and sCD15 (1:200, clone CSLEX1; Biolegend, San Diego, USA) and the EnVision Flex+ kit (K8002, Dako) were used for immunohistochemical staining of the TMA and 769p and 786o cells. Antigen retrieval for CD15 was performed at pH9 and for sCD15 at pH6. All slides were stained with automated immunostainers (autostainer plus, Dako).

Digital Image Analysis

All slides were digitalized using a digital whole slide scanner (NanoZoomer, Hamamatsu Photonics, Hamamatsu, Japan) and analyzed with the HALO® platform (Indica Labs, Corrales, NM, USA). Digital image analysis was performed as previously described [23, 24]. Briefly, the proportion of CD15 and sCD15 positive cells was quantified per TMA core through cell detection algorithms implemented in HALO®. These algorithms can automatically differentiate between stain positive and negative cells after manual adjustment of the thresholds (shown in online suppl. Fig. 1; for all online suppl. material, see https://doi.org/10.1159/000535201). A representative set of cores was used to define the analysis settings and thresholds for each staining. To avoid distortion of results, small areas with strong artificial overlap and detritus were manually excluded from analysis by annotations layers. Missing or erroneous TMA cores were excluded. The results of automated tissue analysis were validated on a set of randomly selected cores.

Cell Culture and Stimulation

The model cell lines for ccRCC 769p and 786o were purchased from ATCC (Rockville, MD). Cell line authentication was performed using Multiplex Cell Authentication by Multiplexion (Heidelberg, Germany). All cell lines were regularly tested for mycoplasma using the MycoAlertTM Mycoplasma Detection Kit (Lonza, Cologne, Germany, LT07-118). The cells were cultured in an RPMI-1640 medium (Thermo Fisher Scientific, Inc.) supplemented with 10% fetal calf serum, 1 mm glutamine, 25 mm glucose, and 1% penicillin-streptomycin (all Thermo Fisher Scientific, Inc.) at 37°C in a humidified atmosphere containing 5% CO2. For CD15 stimulation, cells were transferred to 6-well plates coated with either 20 µg anti-CD15 antibody (clone C3D-1, sc-19648, Santa Cruz, TX, USA) or normal mouse IgM antibody (sc-3881, Santa Cruz, TX, USA) as the negative control. The cells were stimulated for up to 48 h. All experiments were repeated at least three times.

Immunoblotting

Immunoblotting was performed as previously described [25]. Antibodies against ERK1/2 (1:2,000, #9102; Cell Signaling Technology, MA, USA), phospho-ERK1/2 (1:1,000, #9101; Cell Signaling Technology, MA, USA), JNK (1:1,000, #9258; Cell Signaling Technology, MA, USA), phospho-JNK (1:1,000, #4668; Cell Signaling Technology, MA, USA), p38 (1:2,000, #9212; Cell Signaling Technology, MA, USA), and phospho-p38 (1:1,000, #9215; Cell Signaling Technology, MA, USA) were used.

Migration Assay

20,000 769p cells or 30,000 786o cells were seeded into inserts with a defined gap between two separated areas in 6-well plates coated with either anti-CD15 antibody or IgM antibody as the negative control. After 48 h, the inserts were removed and the gap closure was documented photographically every hour for up to 10 h using a Nikon Eclipse TS2. The migratory speed of CD15-stimulated cells in the gap was calculated relative to the cells treated with the control antibody.

Statistical Analysis

For dichotomization of CD15 or sCD15 expression, the Charité Cutoff finder [26] was used to distinguish low- and high-expression levels based on patient survival data. The association between biomarker expression and patients’ survival was calculated using the log rank test and depicted by Kaplan-Meier plots. For multivariate analysis, the Cox regression model was used. Densitometric analysis of the immunoblottings was performed with the open source software ImageJ. Differences between two groups were tested with the paired two-sided t test or Mann-Whitney U test in cases of lack of normal distribution and between three or more groups with Fisher’s exact test for count data. Differences with error probabilities of <0.05 were considered significant.

Patient Collective and Immunohistochemistry

Of the 763 patients in the TMA, 716 were successfully scored for CD15 and 580 for sCD15. The clinicopathological parameters are summarized in Table 1. The median follow-up time was 7.4 years (minimum 0.04 years, maximum 23.7 years, mean 7.9 years). By the end of the follow-up, 248 patients (32.5%) had died from ccRCC. In this subgroup, the median follow-up time was 2.3 years (minimum 0.05 years, maximum 21.3 years, mean 3.6 years). The epithelial cells of the proximal tubules in the tumor-adjacent renal parenchyma were immunohistochemically strongly positive for CD15 and sCD15. In contrast, the tumor cells showed a broad range in the proportion of positively stained tumor cells, particularly for CD15 and to a lesser extent for sCD15 (shown in Fig. 1).

Fig. 1.

Immunohistochemical detection of CD15 and sCD15. The left panel shows (a) CD15 expression in normal kidney parenchyma, (b) low CD15 or (c) high CD15 expression in ccRCC. The right panel shows (d) sCD15 staining in normal kidney parenchyma, (e) low sCD15 expression or (f) high sCD15 expression in ccRCC. Bar indicates 250 µm.

Fig. 1.

Immunohistochemical detection of CD15 and sCD15. The left panel shows (a) CD15 expression in normal kidney parenchyma, (b) low CD15 or (c) high CD15 expression in ccRCC. The right panel shows (d) sCD15 staining in normal kidney parenchyma, (e) low sCD15 expression or (f) high sCD15 expression in ccRCC. Bar indicates 250 µm.

Close modal

Comparison of Biomarker Expression in ccRCC

Unpaired comparison of CD15 and sCD15 expression showed a significant difference in the distribution of CD15 expression (mean 50.6% CD15 positive tumor cells, minimum 0.08%, maximum 99.9%, median of 52.7%) and sCD15 expression (mean 6.5%, minimum 0%, maximum 94.4%, median 0.5%) (shown in Fig. 2a). The paired comparison of both biomarkers revealed subgroups with either low or high expression of both biomarkers and combinations of high CD15 and low sCD15 or vice versa (shown in Fig. 2b).

Fig. 2.

Comparison of CD15 and sCD15 expression. Unpaired (a) and paired (b) comparison of CD15 and sCD15 expression. The p value was calculated using the (paired) Wilcoxon test. Differences with a p value of p < 0.05 were considered significant.

Fig. 2.

Comparison of CD15 and sCD15 expression. Unpaired (a) and paired (b) comparison of CD15 and sCD15 expression. The p value was calculated using the (paired) Wilcoxon test. Differences with a p value of p < 0.05 were considered significant.

Close modal

Prognostic Value of CD15 and sCD15

For dichotomization of the patient cohort, cutoff values were set at 30% CD15 positive and 1% sCD15 positive tumor cells. Accordingly, the patients could be regrouped into CD15 low (n = 268 patients), CD15 high (n = 448 patients), sCD15 low (n = 332 patients), and sCD15 high (n = 242 patients) tumors. The univariate survival analysis revealed a significant association between high CD15 expression and favorable DSS, OS, and PFS (Table 2, shown in Fig. 3). A similar but nonsignificant trend could also be observed for sCD15 (Table 2). Subgroup analysis of the two combined markers confirmed the higher prognostic power of CD15 compared to sCD15 (Table 2). Low CD15 expression in ccRCC correlated significantly with tumor features, indicating aggressive behavior (Table 3). In the multivariate analysis (n = 706 patients), CD15 was not significantly associated with survival (Table 4).

Table 2.

Univariate survival analysis with regard to DSS, OS, and PFS

VariableNDSSOSPFS
HR with 95% CIglobal p valueHR with 95% CIglobal p valueHR with 95% CIglobal p value
sCD15 
 Low 332 Reference  Reference  Reference  
 High 242 0.74 (0.55–1.01) 0.06 0.94 (0.76–1.16) 0.6 0.81 (0.61–1.07) 0.1 
CD15 
 Low 267 Reference  Reference  Reference  
 High 443 0.46 (0.35–0.60) <0.001 0.75 (0.62–0.90) 0.003 0.52 (0.41–0.67) <0.001 
CD15 and sCD15 
 CD15 low, sCD15 low 148 Reference  Reference  Reference  
 CD15 low, sCD15 high 46 0.71 (0.41–1.22) 0.2 0.96 (0.64–1.44) 0.8 0.81 (0.49–1.34) 0.4 
 CD15 high, sCD15 high 188 0.43 (0.3–0.6) <0.001 0.74 (0.57–0.97) 0.03 0.51 (0.36–0.72) <0.001 
 CD15 high, sCD15 low 158 0.39 (0.26–0.58) <0.001 0.7 (0.53–0.94) 0.02 0.48 (0.33–0.69) <0.001 
Grade 
 G1 212 Reference  Reference  Reference  
 G2 408 1.83 (1.27–2.63) 0.001 1.25 (1.00–1.55) 0.049 1.61 (1.17–2.22) 0.003 
 G3 132 6.85 (4.69–10.01) <0.001 3.13 (2.41–4.06) <0.001 5.51 (3.91–7.76) <0.001 
Tumor stage extent (T) 
 T1 412 Reference  Reference  Reference  
 T2 59 3.46 (2.17–5.50) <0.001 1.59 (1.12–2.26) 0.009 3.63 (2.40–5.50) <0.001 
 T3 260 5.12 (3.79–6.91) <0.001 2.33 (1.92–2.83) <0.001 4.73 (3.60–6.22) <0.001 
 T4 25 15.76 (9.55–26.00) <0.001 6.17 (4.02–9.47) <0.001 16.18 (10.08–25.99) <0.001 
Lymph node metastasis (N) 
 No 711 Reference  Reference  Reference  
 Yes 45 5.43 (3.84–7.68) <0.001 3.23 (2.36–4.43) <0.001 5.88 (4.18–8.26) <0.001 
Distant metastasis (M) 
 No 635 Reference  Reference  Reference  
 Yes 121 11.09 (8.50–14.47) <0.001 5.26 (4.20–6.59) <0.001 7.22 (5.61–9.28) <0.001 
Gender 
 Female 292 Reference  Reference  Reference  
 Male 464 1.43 (1.10–1.87) 0.008 1.27 (1.05–1.53) 0.01 1.34 (1.05–1.70) 0.02 
Age at surgery (years) 
 <65 430 Reference  Reference  Reference  
 >65 326 0.97 (0.75–1.25) 0.8 1.77 (1.47–2.13) <0.001 0.94 (0.74–1.19) 0.6 
Karnofsky performance 
 >80% 458 Reference  Reference  Reference  
 <80% 298 1.99 (1.55–2.55) <0.001 2.22 (1.85–2.66) <0.001 1.68 (1.33–2.12) <0.001 
VariableNDSSOSPFS
HR with 95% CIglobal p valueHR with 95% CIglobal p valueHR with 95% CIglobal p value
sCD15 
 Low 332 Reference  Reference  Reference  
 High 242 0.74 (0.55–1.01) 0.06 0.94 (0.76–1.16) 0.6 0.81 (0.61–1.07) 0.1 
CD15 
 Low 267 Reference  Reference  Reference  
 High 443 0.46 (0.35–0.60) <0.001 0.75 (0.62–0.90) 0.003 0.52 (0.41–0.67) <0.001 
CD15 and sCD15 
 CD15 low, sCD15 low 148 Reference  Reference  Reference  
 CD15 low, sCD15 high 46 0.71 (0.41–1.22) 0.2 0.96 (0.64–1.44) 0.8 0.81 (0.49–1.34) 0.4 
 CD15 high, sCD15 high 188 0.43 (0.3–0.6) <0.001 0.74 (0.57–0.97) 0.03 0.51 (0.36–0.72) <0.001 
 CD15 high, sCD15 low 158 0.39 (0.26–0.58) <0.001 0.7 (0.53–0.94) 0.02 0.48 (0.33–0.69) <0.001 
Grade 
 G1 212 Reference  Reference  Reference  
 G2 408 1.83 (1.27–2.63) 0.001 1.25 (1.00–1.55) 0.049 1.61 (1.17–2.22) 0.003 
 G3 132 6.85 (4.69–10.01) <0.001 3.13 (2.41–4.06) <0.001 5.51 (3.91–7.76) <0.001 
Tumor stage extent (T) 
 T1 412 Reference  Reference  Reference  
 T2 59 3.46 (2.17–5.50) <0.001 1.59 (1.12–2.26) 0.009 3.63 (2.40–5.50) <0.001 
 T3 260 5.12 (3.79–6.91) <0.001 2.33 (1.92–2.83) <0.001 4.73 (3.60–6.22) <0.001 
 T4 25 15.76 (9.55–26.00) <0.001 6.17 (4.02–9.47) <0.001 16.18 (10.08–25.99) <0.001 
Lymph node metastasis (N) 
 No 711 Reference  Reference  Reference  
 Yes 45 5.43 (3.84–7.68) <0.001 3.23 (2.36–4.43) <0.001 5.88 (4.18–8.26) <0.001 
Distant metastasis (M) 
 No 635 Reference  Reference  Reference  
 Yes 121 11.09 (8.50–14.47) <0.001 5.26 (4.20–6.59) <0.001 7.22 (5.61–9.28) <0.001 
Gender 
 Female 292 Reference  Reference  Reference  
 Male 464 1.43 (1.10–1.87) 0.008 1.27 (1.05–1.53) 0.01 1.34 (1.05–1.70) 0.02 
Age at surgery (years) 
 <65 430 Reference  Reference  Reference  
 >65 326 0.97 (0.75–1.25) 0.8 1.77 (1.47–2.13) <0.001 0.94 (0.74–1.19) 0.6 
Karnofsky performance 
 >80% 458 Reference  Reference  Reference  
 <80% 298 1.99 (1.55–2.55) <0.001 2.22 (1.85–2.66) <0.001 1.68 (1.33–2.12) <0.001 
Fig. 3.

Survival analysis in ccRCC. Association between (a) OS, (b) DSS, and (c) PFS of patients with ccRCC with low or high CD15 expression represented by Kaplan-Meier plots. The p value was calculated using the log-rank test. Differences with a p value of p < 0.05 were considered significant.

Fig. 3.

Survival analysis in ccRCC. Association between (a) OS, (b) DSS, and (c) PFS of patients with ccRCC with low or high CD15 expression represented by Kaplan-Meier plots. The p value was calculated using the log-rank test. Differences with a p value of p < 0.05 were considered significant.

Close modal
Table 3.

Comparison of CD15 with clinicopathological features

CharacteristicCD15 low, N = 268 (37.4%)*CD15 high, N = 448 (62.6%)*p value
Grade (G), n (%)   <0.001 
 G1 54 (20.1) 153 (34.2)  
 G2 133 (49.6) 257 (57.4)  
 G3 78 (29.1) 37 (8.3)  
 Missing 3 (1.1) 1 (0.2)  
Tumor extent (T), n (%)   <0.001 
 T1 113 (42.2) 290 (64.7)  
 T2 23 (8.6) 35 (7.8)  
 T3 115 (42.9) 116 (25.9)  
 T4 17 (6.3) 7 (1.6)  
Local lymph node metastasis (N), n (%)   0.001 
 N0 243 (90.7) 432 (96.4)  
 N1 25 (9.3) 16 (3.6)  
Distant metastasis (M), n (%)   <0.001 
 M0 202 (75.4) 407 (90.8)  
 M1 66 (24.6) 41 (9.2)  
Sex, n (%)   0.2 
 Female 97 (36.2) 185 (41.3)  
 Male 171 (63.8) 263 (58.7)  
Age at surgery, n (%)   0.5 
 <65 years 158 (59.0) 252 (56.3)  
 ≥65 years 110 (41.0) 196 (43.7)  
Karnofsky performance index, n (%)   0.2 
 >80% 155 (57.8) 283 (63.2)  
 <80% 113 (42.2) 165 (36.8)  
CharacteristicCD15 low, N = 268 (37.4%)*CD15 high, N = 448 (62.6%)*p value
Grade (G), n (%)   <0.001 
 G1 54 (20.1) 153 (34.2)  
 G2 133 (49.6) 257 (57.4)  
 G3 78 (29.1) 37 (8.3)  
 Missing 3 (1.1) 1 (0.2)  
Tumor extent (T), n (%)   <0.001 
 T1 113 (42.2) 290 (64.7)  
 T2 23 (8.6) 35 (7.8)  
 T3 115 (42.9) 116 (25.9)  
 T4 17 (6.3) 7 (1.6)  
Local lymph node metastasis (N), n (%)   0.001 
 N0 243 (90.7) 432 (96.4)  
 N1 25 (9.3) 16 (3.6)  
Distant metastasis (M), n (%)   <0.001 
 M0 202 (75.4) 407 (90.8)  
 M1 66 (24.6) 41 (9.2)  
Sex, n (%)   0.2 
 Female 97 (36.2) 185 (41.3)  
 Male 171 (63.8) 263 (58.7)  
Age at surgery, n (%)   0.5 
 <65 years 158 (59.0) 252 (56.3)  
 ≥65 years 110 (41.0) 196 (43.7)  
Karnofsky performance index, n (%)   0.2 
 >80% 155 (57.8) 283 (63.2)  
 <80% 113 (42.2) 165 (36.8)  

*n (%).

Fisher’s exact test; Pearson’s χ2 test.

Table 4.

Multivariate survival analysis with regard to DSS, OS, and PFS

VariableNDSSOSPFS
HR with 95% CIglobal p valueHR with 95% CIglobal p valueHR with 95% CIglobal p value
CD15 
 Low 264 Reference  Reference  Reference  
 High 442 1.02 (0.76–1.37) 0.9 1.18 (0.96–1.46) 0.1 1.04 (0.79, 1.36) 0.8 
Grade (G) 
 G1 205 Reference  Reference  Reference  
 G2 387 1.03 (0.70–1.51) 0.9 0.92 (0.73–1.16) 0.5 1.05 (0.75, 1.48) 0.8 
 G3 114 2.37 (1.53–3.66) <0.001 1.98 (1.46–2.68) <0.001 2.13 (1.42, 3.18) <0.001 
Tumor stage (T) 
 T1 397 Reference  Reference  Reference  
 T2 55 2.14 (1.30–3.52) 0.003 1.30 (0.90–1.89) 0.2 2.51 (1.61, 3.91) <0.001 
 T3 230 3.40 (2.42–4.77) <0.001 1.76 (1.41–2.20) <0.001 3.32 (2.44, 4.52) <0.001 
 T4 24 2.71 (1.50–4.90) <0.001 2.08 (1.26–3.43) 0.004 3.30 (1.84, 5.90) <0.001 
Lymph node metastasis (N) 
 N0 665 Reference  Reference  Reference  
 N1 41 1.53 (1.02–2.31) 0.04 1.38 (0.96–1.99) 0.09 1.61 (1.07, 2.42) 0.02 
Distant metastasis (M) 
 M0 599 Reference  Reference  Reference  
 M1 107 6.27 (4.55–8.64) <0.001 3.62 (2.77–4.74) <0.001 3.69 (2.69, 5.06) <0.001 
Sex 
 Female 275 Reference  Reference  Reference  
 Male 431 1.17 (0.88–1.56) 0.3 1.27 (1.04–1.55) 0.02 1.12 (0.86, 1.45) 0.4 
Age at surgery (years) 
 <65 403 Reference  Reference  Reference  
 >65 303 1.04 (0.78–1.38) 0.8 1.80 (1.48–2.20) <0.001 0.97 (0.75, 1.26) 0.8 
Karnofsky performance index 
 >80% 431 Reference  Reference  Reference  
 <80% 275 1.55 (1.18–2.04) 0.002 1.81 (1.49–2.20) <0.001 1.34 (1.04, 1.73) 0.02 
VariableNDSSOSPFS
HR with 95% CIglobal p valueHR with 95% CIglobal p valueHR with 95% CIglobal p value
CD15 
 Low 264 Reference  Reference  Reference  
 High 442 1.02 (0.76–1.37) 0.9 1.18 (0.96–1.46) 0.1 1.04 (0.79, 1.36) 0.8 
Grade (G) 
 G1 205 Reference  Reference  Reference  
 G2 387 1.03 (0.70–1.51) 0.9 0.92 (0.73–1.16) 0.5 1.05 (0.75, 1.48) 0.8 
 G3 114 2.37 (1.53–3.66) <0.001 1.98 (1.46–2.68) <0.001 2.13 (1.42, 3.18) <0.001 
Tumor stage (T) 
 T1 397 Reference  Reference  Reference  
 T2 55 2.14 (1.30–3.52) 0.003 1.30 (0.90–1.89) 0.2 2.51 (1.61, 3.91) <0.001 
 T3 230 3.40 (2.42–4.77) <0.001 1.76 (1.41–2.20) <0.001 3.32 (2.44, 4.52) <0.001 
 T4 24 2.71 (1.50–4.90) <0.001 2.08 (1.26–3.43) 0.004 3.30 (1.84, 5.90) <0.001 
Lymph node metastasis (N) 
 N0 665 Reference  Reference  Reference  
 N1 41 1.53 (1.02–2.31) 0.04 1.38 (0.96–1.99) 0.09 1.61 (1.07, 2.42) 0.02 
Distant metastasis (M) 
 M0 599 Reference  Reference  Reference  
 M1 107 6.27 (4.55–8.64) <0.001 3.62 (2.77–4.74) <0.001 3.69 (2.69, 5.06) <0.001 
Sex 
 Female 275 Reference  Reference  Reference  
 Male 431 1.17 (0.88–1.56) 0.3 1.27 (1.04–1.55) 0.02 1.12 (0.86, 1.45) 0.4 
Age at surgery (years) 
 <65 403 Reference  Reference  Reference  
 >65 303 1.04 (0.78–1.38) 0.8 1.80 (1.48–2.20) <0.001 0.97 (0.75, 1.26) 0.8 
Karnofsky performance index 
 >80% 431 Reference  Reference  Reference  
 <80% 275 1.55 (1.18–2.04) 0.002 1.81 (1.49–2.20) <0.001 1.34 (1.04, 1.73) 0.02 

CD15 Expression and Metastasis

To investigate a putative role of CD15 in the formation of metastasis, CD15 positivity of nonmetastasized ccRCC was compared with tumors that had either lymphogenic spread or synchronous or metachronous metastasis (shown in Fig. 4). On average, 55.1% ± 36.4 (mean ± standard deviation) of tumor cells in ccRCC without metastatic spread were CD15 positive compared to 44.0% ± 38.5 in tumors with lymph node metastasis and 35.4% ± 37.6 in tumors with synchronous metastasis. The latter was significantly lower compared to nonmetastasized tumors (p ≤ 0.001). 50.5% ± 38.3 of tumor cells in tumors with metachronous metastasis showed CD15 positivity and thus no significant difference to nonmetastasized tumors. Since metastasis is not only dependent on CD15 expression alone and CD15 correlated significantly with tumor grading (Table 3), the proportion of ccRCC with poor grading was compared in the groups of tumors with and without synchronous metastasis. 43.2% of tumors with synchronous metastasis were poorly differentiated compared to 12.6% of nonmetastasized tumors (p < 0.001).

Fig. 4.

Comparison of CD15 expression of nonmetastasized ccRCC and ccRCC with metastasis. ccRCC were grouped into tumors without any metastasis (N0/M0) and tumors with lymph node metastasis (N1), synchronous (prim. M1), or metachronous distant metastasis (sec. M1). The range of the CD15 expression in the particular groups was plotted. The p value was calculated using the Fisher’s exact test. Differences with a p value of p < 0.05 were considered significant. **p ≤ 0.01; ***p ≤ 0.001.

Fig. 4.

Comparison of CD15 expression of nonmetastasized ccRCC and ccRCC with metastasis. ccRCC were grouped into tumors without any metastasis (N0/M0) and tumors with lymph node metastasis (N1), synchronous (prim. M1), or metachronous distant metastasis (sec. M1). The range of the CD15 expression in the particular groups was plotted. The p value was calculated using the Fisher’s exact test. Differences with a p value of p < 0.05 were considered significant. **p ≤ 0.01; ***p ≤ 0.001.

Close modal

Effects of CD15 Stimulation on the Migration of ccRCC Cell Lines

For functional analyses, the CD15 positive 769p cells (shown in Fig. 5a) and CD15 negative 786o cells (shown in Fig. 5b) were used. To investigate the effect of CD15 on cellular motility, 769p and 786o cells were stimulated with an anti-CD15-antibody or control antibody for 48 h, and the migration into a defined gap was quantified. CD15-stimulated 769p cells (p = 0.02) and 786o cells (p = 0.04) migrated significantly slower compared to control (shown in Fig. 5c, d). Since CD15-mediated stimulation of Hodgkin lymphoma cells and monocytes resulted in an activation of activating protein 1 (AP-1) (12,13), the impact of CD15 on AP-1 activating pathways ERK1/2, JNK, p38 was examined. Immunoblotting after cell stimulation for 48 h revealed a slight, yet significant, increase of p38 in 786o cells (p = 0.03). However, there was no effect on the phosphorylated protein pp38 (p = 0.99). This effect could not be detected in 769p cells (p38: p = 0.1; pp38: p = 0.3). The same was true for ERK1/2 (769p and 786o: p = 0.3) and JNK (769p: p = 0.3; 786o: p = 0.99) and the respective phosphorylation levels of pERK1/2 (769p: p = 0.3; 786o: p = 0.99) and pJNK (769p: p = 0.3; 786o: p = 0.99). Representative immunoblots are shown in Figure 5e and f.

Fig. 5.

In vitro investigations on the effect of CD15 stimulation on the ccRCC model cell lines 769p and 786o. a 769p cells and (b) 786o cells were immunocytochemically stained against CD15. After stimulation with CD15 or control antibody for 48 h (c–f), the migration velocity of (c) 769p cells or (d) 786o cells was determined in comparison to the control. Representative immunoblots of ERK1/2 and phospho ERK1/2 (pERK1/2), c-Jun (JNK), and phospho c-Jun (pJNK), p38 and phospho p38 (pp38) of (e) 769p cells or (f) 786o cells are shown. The p value was calculated using the paired t test. Differences with a p value of p < 0.05 were considered significant.

Fig. 5.

In vitro investigations on the effect of CD15 stimulation on the ccRCC model cell lines 769p and 786o. a 769p cells and (b) 786o cells were immunocytochemically stained against CD15. After stimulation with CD15 or control antibody for 48 h (c–f), the migration velocity of (c) 769p cells or (d) 786o cells was determined in comparison to the control. Representative immunoblots of ERK1/2 and phospho ERK1/2 (pERK1/2), c-Jun (JNK), and phospho c-Jun (pJNK), p38 and phospho p38 (pp38) of (e) 769p cells or (f) 786o cells are shown. The p value was calculated using the paired t test. Differences with a p value of p < 0.05 were considered significant.

Close modal

CD15 and sCD15 are expressed on the cell surface of non-neoplastic and neoplastic cells, respectively. They are involved in cell adhesion and may contribute to tumor cell metastasis. Thus, the aim of this study was to examine CD15 and sCD15 expression in ccRCC with regard to their prognostic value and contribution to metastasis.

Immunohistochemistry showed that CD15 and sCD15 are expressed in the proximal tubules of the kidney parenchyma and ccRCC. The comparison of their expression in ccRCC revealed a stronger and more variable expression of CD15 than that of sCD15 (shown in Fig. 2). This strengthens the hypothesis of ccRCC arising from the cells of the proximal convoluted tubule [27]. However, the difference between the expression profiles of CD15 and sCD15 indicates a different biological significance in ccRCC. In our study, sCD15 had no significant prognostic value (Table 2), whereas other studies found sCD15 as a marker of adverse outcome in RCC [16‒18]. However, different antibodies and scoring systems were used in these studies [16‒18]. The assessment of sCD15 in other neoplasias remains controversial. sCD15 was associated with an unfavorable outcome in nonsmall cellular lung carcinoma, colon carcinoma, or breast carcinoma [28‒30], but with a favorable outcome in endometrial cancer [31]. Moreover, it needs to be considered that sCD15 is coexpressed with tumor-specific biomarkers, e.g., carcinoembryonic antigen in gastric cancer [32] or MUC1 in bladder cancer [33]. These coexpressions might explain the variable prognostic value. So far, colocalizations of sCD15 with other surface molecules have not been examined in RCC.

In our study, the prognostic value of CD15 was superior in the survival analyses compared to that of sCD15, so we focused on CD15 for further investigations. High CD15 expression was associated with a favorable outcome in our ccRCC study cohort including cases with advanced tumor stages (Table 2, shown in Fig. 3). This is in line with the results of previously published data investigating localized ccRCC only [15]. In hepatocellular carcinoma, CD15-positive tumors had more intrahepatic metastases compared to CD15-negative tumors [34]. In medullary thyroid carcinoma, patients with CD15-positive tumors had a higher tumor load and positive lymph node metastases as well as an adverse outcome after tumor recurrence [35]. In nonsmall cellular lung carcinoma, CD15 was associated with a decreased performance status and a biomarker of shortened survival time [28]. According to the results in ccRCC, a high CD15 expression was associated with better survival in squamous cell carcinomas of the head and neck [36]. The tumor cells in Hodgkin’s lymphoma are known to express CD15 [37] and its expression, together with the lack of sCD15, is associated with a favorable outcome [38, 39]. In summary, the prognostic value of CD15 depends on the tumor entity. In ccRCC, it is primarily an indicator of grading and thus biological aggressiveness. This relationship most likely explains the lacking significance of CD15 in the multivariate survival analysis (Table 4).

CD15 is thought to contribute to cell adhesion and consequently to metastasis. In our study, nonmetastasized ccRCC had a significantly higher CD15 expression compared to tumors with synchronous metastasis (shown in Fig. 4). The latter subset contained a higher proportion of poorly differentiated tumors compared to nonmetastasized ccRCC. The CD15 expression of metachronously metastasized ccRCC was comparable to nonmetastasized tumors. This connection leads to the hypothesis of a contribution of CD15 to metachronous metastasis in ccRCC. A putative link is the increased activation of AP-1, a pathway associated with invasiveness, after crosslinking of CD15, as was shown on monocytes or Hodgkin lymphoma cells [12, 13]. Moreover, investigations on ccRCC with venous tumor thrombi demonstrated that AP-1 was upregulated in invasive thrombi compared to the primary tumor [40]. Thus, we also tested the impact of CD15 stimulation on the AP-1 activating pathways ERK, JNK, and p38, but found a minor increase of p38 in 786o cells only. The level of the activated phosphorylated protein pp38 was not altered. There were no significant changes in the other examined pathways on both cell lines. Thus, there is no evidence for AP-1 activation after CD15 stimulation in ccRCC. However, both cell lines, and more pronounced 769p cells, migrated significantly slower after CD15 treatment compared to control (shown in Fig. 5).

In summary, the direct comparison of CD15 and sCD15 in ccRCC showed a strong risk prediction of CD15 but not of sCD15. Including the CD15 status of ccRCC in the pathology report could help estimate the biologic aggressiveness of ccRCC and to guide reasonable decisions on further therapy for the patients. CD15 is an interesting novel target in ccRCC since CD15 stimulation exerted antitumor effects in vitro. The data show no involvement of AP-1. Further research deciphering the underlying signaling pathways and the putative contribution to metachronous metastasis is needed.

We thank Jutta Richter, Silke Mitschke, and Bonny Adami for excellent technical assistance. This project was supported by the Tissue Bank of the University Medical Center Mainz.

This study protocol was reviewed and approved by the Ethics Committee of the University of Heidelberg, approval number 206-2005. The need for informed consent was waived by the Ethics Committee of the University of Heidelberg.

The authors have no conflicts of interest to declare.

No funding was received for this study.

Philipp Joachim Stenzel: conceptualization, methodology, formal analysis, investigation, writing – original draft, visualization, and project administration; Mario Schindeldecker: methodology and formal analysis; Larissa Seidmann, Esther Herpel, Markus Hohenfellner, Gencay Hatiboglu, Sebastian Foersch, and Stefan Porubsky: writing – review and editing; Stephan Macher-Goeppinger: conceptualization, writing – review and editing, and supervision; Wilfried Roth: conceptualization, methodology, resources, writing – review and editing; and supervision; Katrin Elisabeth Tagscherer: conceptualization, methodology, investigation, resources, writing – review and editing, and supervision.

All data generated or analyzed during this study are included in this article and its online supplementary material. Further inquiries can be directed to the corresponding author.

1.
Choueiri TK, Motzer RJ. Systemic therapy for metastatic renal-cell carcinoma. N Engl J Med. 2017;376(4):354–66.
2.
Graves A, Hessamodini H, Wong G, Lim WH. Metastatic renal cell carcinoma: update on epidemiology, genetics, and therapeutic modalities. Immunotargets Ther. 2013;2:73–90.
3.
Läubli H, Borsig L. Selectins promote tumor metastasis. Semin Cancer Biol. 2010;20(3):169–77.
4.
St Hill CA. Interactions between endothelial selectins and cancer cells regulate metastasis. Fron Biosci. 2011;16(9):3233–51.
5.
Li J, King MR. Adhesion receptors as therapeutic targets for circulating tumor cells. Front Oncol. 2012;2(79):79.
6.
Kannagi R, Izawa M, Koike T, Miyazaki K, Kimura N. Carbohydrate-mediated cell adhesion in cancer metastasis and angiogenesis. Cancer Sci. 2004;95(5):377–84.
7.
Skacel PO, Edwards AJ, Harrison CT, Watkins WM. Enzymic control of the expression of the X determinant (CD15) in human myeloid cells during maturation: the regulatory role of 6-sialytransferase. Blood. 1991;78(6):1452–60.
8.
Nakayama F, Nishihara S, Iwasaki H, Kudo T, Okubo R, Kaneko M, et al. CD15 expression in mature granulocytes is determined by alpha 1,3-fucosyltransferase IX, but in promyelocytes and monocytes by alpha 1,3-fucosyltransferase IV. J Biol Chem. 2001;276(19):16100–6.
9.
Gadhoum SZ, Sackstein R. CD15 expression in human myeloid cell differentiation is regulated by sialidase activity. Nat Chem Biol. 2008;4(12):751–7.
10.
Croce MV, Isla-Larrain M, Rabassa ME, Demichelis S, Colussi AG, Crespo M, et al. Lewis x is highly expressed in normal tissues: a comparative immunohistochemical study and literature revision. Pathol Oncol Res. 2007;13(2):130–8.
11.
Kerr M, Stocks S. The role of CD15-(LeX)-related carbohydrates in neutrophil adhesion. Histochem J. 1992;24(11):811–26.
12.
Lo S, Golenbock D, Sass P, Maskati A, Xu H, Silverstein R. Engagement of the Lewis X antigen (CD15) results in monocyte activation. Blood. 1997;89(1):307–14.
13.
Ohana-Malka O, Benharroch D, Isakov N, Prinsloo I, Shubinsky G, Sacks M, et al. Selectins and anti-CD15 (Lewis x/a) antibodies transmit activation signals in Hodgkin’s lymphoma-derived cell lines. Exp Hematol. 2003;31(11):1057–65.
14.
López JI, Moreno V, García H, Antón I, Robles A, Oñate JM, et al. Renal cell carcinoma in young adults: a study of 130 cases and a review of previous series. Urol Int. 2010;84(3):292–300.
15.
Wu CY, Huo JP, Zhang XK, Zhang YJ, Hu WM, Yang P, et al. Loss of CD15 expression in clear cell renal cell carcinoma is correlated with worse prognosis in Chinese patients. Jpn J Clin Oncol. 2017;47(12):1182–8.
16.
Tozawa K, Okamoto T, Kawai N, Hashimoto Y, Hayashi Y, Kohri K. Positive correlation between sialyl Lewis X expression and pathologic findings in renal cell carcinoma. Kidney Int. 2005;67(4):1391–6.
17.
Kobayashi M, Morita T. Significant expression patterns of Lewis x-related antigens as a prognostic predictor of low-stage renal cell carcinomas. Anticancer Res. 2010;30(2):593–9.
18.
Tanio M, Muramoto A, Hoshino H, Murahashi M, Imamura Y, Yokoyama O, et al. Expression of functional E-selectin ligands on the plasma membrane of carcinoma cells correlates with poor prognosis in clear cell renal cell carcinoma. Urol Oncol. 2021;39(5):302.e9–18.
19.
Macher-Goeppinger S, Aulmann S, Wagener N, Funke B, Tagscherer KE, Haferkamp A, et al. Decoy receptor 3 is a prognostic factor in renal cell cancer. Neoplasia. 2008;10(10):1049–56.
20.
Macher-Goeppinger S, Aulmann S, Tagscherer KE, Wagener N, Haferkamp A, Penzel R, et al. Prognostic value of tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) and TRAIL receptors in renal cell cancer. Clin Cancer Res. 2009;15(2):650–9.
21.
Donskov F, Xie W, Overby A, Wells JC, Fraccon AP, Sacco CS, et al. Synchronous versus metachronous metastatic disease: impact of time to metastasis on patient outcome-results from the international metastatic renal cell carcinoma database consortium. Eur Urol Oncol. 2020;3(4):530–9.
22.
Sobin L, Gospodarowicz M, Wittekind C. TNM classification of malignant tumours. 7th edn. 2009.
23.
Foersch S, Schindeldecker M, Keith M, Tagscherer KE, Fernandez A, Stenzel PJ, et al. Prognostic relevance of androgen receptor expression in renal cell carcinomas. Oncotarget. 2017;8(45):78545–55.
24.
Stenzel PJ, Schindeldecker M, Tagscherer KE, Foersch S, Herpel E, Hohenfellner M, et al. Prognostic and predictive value of tumor-infiltrating leukocytes and of immune checkpoint molecules PD1 and PDL1 in clear cell renal cell carcinoma. Transl Oncol. 2020;13(2):336–45.
25.
Macher-Goeppinger S, Keith M, Hatiboglu G, Hohenfellner M, Schirmacher P, Roth W, et al. Expression and functional characterization of the BNIP3 protein in renal cell carcinomas. Transl Oncol. 2017;10(6):869–75.
26.
Budczies J, Klauschen F, Sinn BV, Gyorffy B, Schmitt WD, Darb-Esfahani S, et al. Cutoff finder: a comprehensive and straightforward web application enabling rapid biomarker cutoff optimization. PLoS One. 2012;7(12):e51862.
27.
Cairns P. Renal cell carcinoma. Cancer Biomark. 2011;9(1–6):461–73.
28.
Kadota A, Masutani M, Takei M, Horie T. Evaluation of expression of CD15 and sCD15 in non-small cell lung cancer. Int J Oncol. 1999;15(6):1081–9.
29.
Grabowski P, Mann B, Mansmann U, Lövin N, Foss HD, Berger G, et al. Expression of sialyl-Le(x) antigen defined by MAb AM-3 is an independent prognostic marker in colorectal carcinoma patients. Int J Cancer. 2000;88(2):281–6.
30.
Song Y, Sun H, Wu K, Lyu J, Zhang J, Gu F, et al. sLex expression in invasive micropapillary breast carcinoma is associated with poor prognosis and can be combined with MUC1/EMA as a supplementary diagnostic indicator. Cancer Biol Med. 2021;18(2):477–89.
31.
Kolben T, Müller L, Meister S, Keilmann L, Buschmann C, Trillsch F, et al. Blood group antigens SLeX, SLeA, and LeY as prognostic markers in endometrial cancer. J Cancer Res Clin Oncol. 2022;148(12):3323–35.
32.
Gomes C, Almeida A, Barreira A, Calheiros J, Pinto F, Abrantes R, et al. Carcinoembryonic antigen carrying SLeX as a new biomarker of more aggressive gastric carcinomas. Theranostics. 2019;9(24):7431–46.
33.
Shinagawa T, Hoshino H, Taga M, Sakai Y, Imamura Y, Yokoyama O, et al. Clinicopathological implications to micropapillary bladder urothelial carcinoma of the presence of sialyl Lewis X-decorated mucin 1 in stroma-facing membranes. Urol Oncol. 2017;35(10):606.e17–23.
34.
Torii A, Nakayama A, Harada A, Nakao A, Nonami T, Sakamoto J, et al. Expression of the CD15 antigen in hepatocellular carcinoma. Cancer. 1993;71(12):3864–7.
35.
Neuhold N, Längle F, Gnant M, Hollenstein U, Niederle B. Relationship of CD15 immunoreactivity and prognosis in sporadic medullary thyroid carcinoma. J Cancer Res Clin Oncol. 1992;118(8):629–34.
36.
Rabassa ME, Pereyra A, Pereyra L, Segal-Eiras A, Abba MC, Croce MV. Lewis x antigen is associated to head and neck squamous cell carcinoma survival. Pathol Oncol Res. 2018;24(3):525–31.
37.
Zukerberg L, Collins A, Ferry J, Harris N. Coexpression of CD15 and CD20 by Reed-Sternberg cells in Hodgkin’s disease. Am J Pathol. 1991;139(3):475–83.
38.
Benharroch D, Dima E, Levy A, Ohana-Malka O, Ariad S, Prinsloo I, et al. Differential expression of sialyl and non-sialyl-CD15 antigens on Hodgkin-Reed-Sternberg cells: significance in Hodgkin’s disease. Leuk Lymphoma. 2000;39(1–2):185–94.
39.
Benharroch D, Pilosof S, Gopas J, Levi I. Primary refractory and relapsed classical Hodgkin lymphoma: significance of differential CD15 expression in hodgkin-reed-sternberg cells. J Cancer. 2012;3:322–7.
40.
Kim K, Zhou Q, Christie A, Stevens C, Ma Y, Onabolu O, et al. Determinants of renal cell carcinoma invasion and metastatic competence. Nat Commun. 2021;12(1):5760.