Int J Clin Oncol DOI 10.1007/s10147-014-0710-6
ORIGINAL ARTICLE
Combination of MUC1 and MUC4 expression predicts clinical outcome in patients with oral squamous cell carcinoma Yoshiaki Kamikawa • Yuji Kanmura • Tomofumi Hamada • Norishige Yamada • Muzafar A. Macha • Surinder K. Batra • Michiyo Higashi • Suguru Yonezawa • Kazumasa Sugihara
Received: 19 February 2014 / Accepted: 12 May 2014 Ó Japan Society of Clinical Oncology 2014
Abstract Background Both MUC1 and MUC4 are high molecular weight glycoproteins and are independent indicators of worse prognosis in many human epithelial cancers including oral squamous cell carcinoma (OSCC). However, there has been no investigation of the clinical importance of the co-expression of MUC1 and MUC4 in OSCC. The aim of this study was to evaluate the co-expression profile of MUC1/MUC4 and analyze the prognostic significance in OSCC. Methods We examined the expression profile of MUC1 and MUC4 in OSCC tissues from 206 patients using immunohistochemistry. The co-expression profile of MUC1/MUC4 and its prognostic significance in OSCC was statistically analyzed.
Y. Kamikawa and Y. Kanmura contributed equally to this work.
Results MUC1 and MUC4 overexpression were strongly correlated with each other (p \ 0.0001) and a combination of both MUC1 and MUC4 expression was a powerful indicator for tumor aggressiveness such as tumor size (p = 0.014), lymph node metastasis (0.0001), tumor stage (p = 0.006), diffuse invasion (p = 0.028), and vascular invasion (p = 0.014). The MUC1/MUC4 double-positive patients showed the poorest overall and disease-free survival. Multivariate analysis revealed that MUC1/MUC4 double-positivity was the strong independent prognostic factor for overall and disease-free survival (p = 0.007 and (p = 0.0019), in addition to regional recurrence (p = 0.0025). Conclusions Taken together, these observations indicate that the use of a combination of MUC1/MUC4 can predict outcomes for patients with OSCC. This combination is also a useful marker for predicting regional recurrence. MUC1 and MUC4 may be attractive targets for the selection of treatment methods in OSCC.
Electronic supplementary material The online version of this article (doi:10.1007/s10147-014-0710-6) contains supplementary material, which is available to authorized users. Y. Kamikawa T. Hamada (&) K. Sugihara Department of Oral Surgery, Kagoshima University Hospital, 8-35-1 Sakuragaoka, Kagoshima 890-8544, Japan e-mail: [email protected]
M. A. Macha S. K. Batra Department of Biochemistry and Molecular Biology, Buffett Cancer Center, Eppley Cancer Institute, University of Nebraska Medical Center, 985870 Nebraska Medical Center, Omaha, NE 68198-5870, USA
Y. Kanmura K. Sugihara Department of Maxillofacial Diagnostic and Surgical Science, Field of Oral and Maxillofacial Rehabilitation, Kagoshima University Graduate School of Medical and Dental Sciences, 8-35-1 Sakuragaoka, Kagoshima 890-8544, Japan
M. Higashi S. Yonezawa Department of Human Pathology, Field of Oncology, Kagoshima University Graduate School of Medical and Dental Sciences, 8-35-1 Sakuragaoka, Kagoshima 890-8544, Japan
N. Yamada Division of Reproductive Sciences, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave., Cincinnati, Ohio 45229, USA
123
Int J Clin Oncol
Keywords Oral squamous cell carcinoma MUC1 MUC4 Immunohistochemistry Prognosis
Materials and methods Patients
Introduction Oral cancer is a serious problem in many parts of the world, with approximately 263,900 new cases and 128,000 deaths worldwide in 2008, and oral squamous cell carcinomas (OSCCs) comprise [90 % of these cases [1]. Survival rates of OSCC have not improved significantly over the past 30 years, despite outstanding progress in clinical and basic research [2]. One of the main prognosticators of the survival of patients with OSCC is regional lymph node metastasis, but such metastasis is quite often subclinical or occult at the time of diagnosis [3, 4]. Additionally, patients treated with curative surgery often have poor outcomes because of unexpected events such as regional recurrence [3, 5]. Therefore prognostic factors that predict clinical outcomes may be helpful in guiding treatment decisions. However, despite the substantial research effort made over decades, very few prognostic markers and predictive assays have been established in routine clinical diagnosis and treatment of OSCC [6]. Mucins are high molecular weight glycoproteins that help to protect and lubricate luminal surfaces under normal physiological conditions [7, 8]. To date, 21 different mucins have been found in humans and functionally classified into secreted and membrane-bound types. These mucins are involved in cell differentiation, renewal of the epithelium and modulation of cell adhesion, immune response, and cell signaling [7–9]. Our previous studies have reported the deregulation of both MUC1 and MUC4 in several human malignancies and these are associated with aggressive tumor behavior and poor outcomes [10–29]. In OSCC, aberrant expression of MUC1 [30] or MUC4 [31] is an independent prognostic factor indicating a poor prognosis. Interestingly, previous studies of intrahepatic cholangiocarcinoma-mass forming type [10] and gallbladder carcinoma [32] demonstrated a significant correlation between MUC1 and MUC4 expression, although these mucins are encoded by different genes (MUC1, chromosome 1q21; MUC4, chromosome 3q29) and are involved in different signaling pathways [7– 9]. However, co-expression status and its clinical significance of MUC1 and MUC4 have not been investigated in OSCCs. Therefore, in the present study, we examined the expression profile of MUC1 and MUC4 and found their coexpression in OSCC tissues to be significant. Furthermore, MUC1/MUC4 co-expression was correlated with tumor aggressiveness, and MUC1/MUC4 double-positivity was observed to be an independent worse prognostic factor with decreased overall and disease-free survival, in addition to regional recurrence in patients with OSCC.
123
In the present study, we used the patient series collected for our previous studies of MUC1 [30] and MUC4 [31] expression in OSCC. The information of vital status of all patients was updated and survival rates were re-calculated. This study was conducted in accordance with the guiding principles of the Declaration of Helsinki. After ethics approval from the Human Ethics Committee of Kagoshima University Graduate School of Medical and Dental Sciences (permission no. 34, 65, 99, 101), a retrospective search was carried out to identify patients pathologically diagnosed with OSCC. This study included 206 patients who underwent surgical resection for primary oral cancer between 1992 and 2008 at the Department of Oral Surgery, Kagoshima University Hospital in Kagoshima, Japan. Seventy-eight out of 206 (38 %) patients received neoadjuvant chemotherapy or postoperative chemotherapy that used fluorouracil or platinum-containing anti-cancer drugs such as CDDP and CBDCA. There were 128 out of 206 (62 %) patients that were initially treated with only surgical resection. Patients were excluded if they had received previous treatment for oral cancer, had distant metastasis at the time of their first visit, or did not undergo surgical treatment. Informed consent was obtained to use surgical specimens for this study from all of the patients. Histopathological analysis All specimens were fixed in formalin, embedded in paraffin, and cut into 4-lm-thick sections for hematoxylin and eosin (HE) staining for histopathological analysis and for immunohistochemical analysis of MUC1 and MUC4. Detailed tumor characteristics, including the tumor-nodemetastasis (TNM) classification, differentiation, mode of invasion, vascular invasion, lymphatic permeation, and perineural infiltration were evaluated histologically. Anatomical sites, tumor differentiation, TNM classification, and stage grouping of the tumor were classified according to the TNM classification of malignant tumors [33]. Aggressiveness of invasion was assessed by classification of the mode of invasion [34]. Immunohistochemistry Immunohistochemical staining was performed as described previously [30, 31]. Briefly, for MUC1 staining, we used a mouse monoclonal antibody DF3 (Toray-Fuji Bionics, Tokyo, Japan) and MUC4 was detected by a mouse monoclonal antibody8G7 (generated by one of us, S. K. Batra). Antigen retrieval was performed using CC1
Int J Clin Oncol
antigen retrieval buffer (Ventana Medical Systems, Tucson, AZ, USA) for MUC1 and by 0.01 M citrate buffer (pH 6.0) in water bath at 80 °C for 20 min for MUC4. After washing with PBS, tissue sections were incubated at 37 °C for 32 min for DF3 (diluted at 1:10) [30] and at 4 °C for 16 h for 8G7 (diluted at 1:3,000) [31]. Primary antibody was replaced by hybridoma culture medium, normal mouse serum, or PBS in negative controls. Two blinded investigators (Y. K. and T. H.) independently evaluated the results of immunostaining without knowledge of the patient outcomes. According to preliminary statistical analysis by receiver operating characteristic (ROC) curve, specimens showing more than 5 % positively stained carcinoma cells were graded as positive [30, 31]. Statistical analysis Associations between mucin expression and clinicopathological characteristics were evaluated by the Chi-square test. The accuracy of the combination of MUC1/MUC4 positivity was evaluated from the ROC curve analysis by calculating the area under the curve (AUC). A ROC curve was constructed by plotting the sensitivity (true-positive rate) against 1-specificity (false-positive rate) for detection of tumor death and recurrence. Overall and disease-free survival of the patients was compared between groups according to the Kaplan–Meier method. Differences in survival rates were evaluated using the log-rank test. Overall survival was measured from the date of surgery to the date of the last follow-up examination or death, whereas disease-free survival was measured from the date of surgery to the date that local recurrence and/or regional recurrence were first detected. We defined ‘‘regional recurrence’’ as regional lymph node metastasis that is detected after initial treatment in all patients regardless of neck dissection. For analysis of prognostic factors, six tumor locations were classified into one of two groups: the better survival group (lip, buccal, tongue, and mandible, n = 157) and the poor survival group (maxilla and floor of the mouth, n = 49) according to preliminary survival analyses described previously [31]. Univariate and multivariate analyses of overall and disease-free survival were performed using the Cox proportional hazards regression model. Variables that showed p \ 0.10 in the univariate analysis in a previous study [30] were used to create a multivariate model in this study to determine whether mucin expression was an independent prognostic factor. In multivariate analysis, the tumor stage was excluded for the model, because the tumor stage contains both T and N classifications, which would negate the statistical significance of each other. In addition to survival analysis, we performed univariate and multivariate analyses of the relative risk for regional recurrence.
Statistical analysis was performed using JMP software (version 9.0.2; SAS Institute, Cary, NC, USA). All statistics were two-sided, and a probability of p \ 0.05 was considered to be significant.
Results MUC1 and MUC4 expression and their correlation in OSCC Expression of MUC1 and MUC4 was analyzed in OSCC tissues by IHC and representative patterns of combination of expression with HE staining as reference are illustrated in Fig. 1. The statistical correlation between MUC1 and MUC4 expressions in 206 OSCC patients is shown in Table 1. There were 95/206 (46 %) MUC1-positive and 80/206 (39 %) MUC4-positive patients. Double-positivity of MUC1 and MUC4 was observed in 60 of 206 (29 %) OSCC cases (Fig. 1a), whereas double negativity was found in 91 (44 %) patients (Fig. 1d). Taken together, MUC1 and MUC4 expression was strongly correlated with each other (p \ 0.0001, Table 1). Combinations of MUC1 and MUC4 expression in OSCC To determine the most appropriate combination of MUC1 and MUC4 expression as a prognostic factor in OSCC, we performed ROC analysis and calculated the AUC (Table 2). When patients were divided into three groups (MUC1/MUC4 double-positive (?/?) group (Fig. 1a), single-positive (?/- or -/?) group (Fig. 1b, c), and double-negative (-/-) group (Fig. 1d), the AUC for tumor death was 0.658 and that for recurrence was 0.668. These AUCs were higher than those of MUC1 or MUC4 single analysis (0.625 and 0.616 for MUC1, 0.614 and 0.639 for MUC4, respectively), suggesting that the combination of MUC1 and MUC4 is a more useful prognostic factor than single evaluation to predict the outcomes of patients with OSCC. Therefore, we used ‘‘MUC1/MUC4 double-positive vs. single-positive vs. double-negative’’ as the grouping method for further analysis. Association between MUC1/MUC4 co-expression and clinicopathologic characteristics Associations between MUC1/MUC4 co-expression and clinicopathological characteristics are summarized in Table 3, and factors with statistical significance are illustrated in Fig. 2. The percentages of patients with a large tumor size (p = 0.014), positive nodal metastasis (p = 0.0001), advanced tumor stage (p = 0.006), diffuse invasion
123
Int J Clin Oncol Fig. 1 Representative patterns of combination of MUC1/ MUC4 immunohistochemistry. a Double-positive group (MUC1?/MUC4?, n = 60 of total 206 patients). Both MUC1 and MUC4 showed strong positivity in invasive carcinoma cells just beneath epithelium. MUC4 was focally positive in dysplastic epithelium in this case. b Single-positive group (MUC1?/MUC4-, n = 35). Tumor cells showed MUC1positive but MUC4-negative expression in this case. c Singlepositive group (MUC1-/ MUC4?, n = 20). c Contrary to b, tumor cells showed MUC1negative but MUC4-positive expression in this case. d Double negative group (MUC1-/MUC4-, n = 91). The tumor cells were negative for both MUC1 and MUC4
Table 1 Correlation (n = 206) MUC1 expression
Negative
between
MUC1
MUC4 expression Negative
Positive
91
20
and
MUC4
expression
Total
Table 2 The accuracy from combination of MUC1 and MUC4 Grouping method
Tumor death
Local or neck recurrence
0.625
0.616
0.614
0.639
Double-positive vs. other
0.586
0.612
Double-negative vs. other
0.654
0.643
Double-negative vs. single-positive vs. double-positive
0.658
0.668
111
Positive
35
60
95
Total
126
80
206 p \ 0.0001 (Chi-square)
MUC1 Positive vs. negative MUC4 Positive vs. negative
of cancer cells (p = 0.028), vascular invasion (p = 0.014), tumor death (p = 0.005), and regional recurrence (p = 0.0025) were significantly correlated with MUC1/ MUC4 co-expression. As shown in Fig. 2, tumor aggressiveness was strongly associated with MUC1/MUC4 double-positivity rather than single-positivity. Overall, the use of MUC1/MUC4 combination was a powerful indicator for tumor aggressiveness and poor outcome in OSCC. There was no significant difference in the proportion of patients who received chemotherapy between the MUC1-positive group and the MUC1-negative group and between the MUC4-positive group and the MUC4-negative group. Also, there was no significant difference in the type of anti-cancer drug between these groups.
123
AUC in ROC analysis endpoint
MUC1/MUC4
AUC area under the curve, ROC receiver operating characteristic
Clinical significance of MUC1/MUC4 combination for overall survival The 5-year overall survival rate of all patients was 79.4 %. The 5-year overall survival rate of the MUC1/MUC4 doublenegative group was obviously high (92.2 %) and significantly better than that of the double-positive group (72.0 %) and the
Int J Clin Oncol Table 3 Association between MUC1/MUC4 co-expression and clinicopathologic factors Variable
MUC1/MUC4 expression no. of patients (%) Doublenegative n = 86
Singlepositive n = 55
Table 3 continued Variable
p
MUC1/MUC4 expression no. of patients (%) Doublenegative n = 86
Singlepositive n = 55
Doublepositive n = 65
52 (95)
63 (97)
3 (5)
2 (3)
75 (87)
39 (71)
45 (69)
Death from OSCC
6 (7)
13 (24)
19 (29)
Death from other than OSCC
5 (6)
3 (5)
1 (2)
Doublepositive n = 65
Perineural infiltration Absent 85 (99)
Age (years) \65
38 (44)
28 (51)
25 (38)
C65
48 (56)
27 (49)
40 (62)
30 (35)
17 (31)
26 (40)
56 (65)
38 (69)
39 (60)
Tongue
35 (41)
26 (47)
31 (48)
Mandible
16 (19)
13 (24)
13 (20)
Absent
80 (93)
43 (78)
47 (72)
Maxilla
6 (19)
5 (9)
9 (14)
Present
6 (7)
12 (22)
18 (28)
Buccal
0.39
Present
Male Tumor location
Alive 0.58
0.47
11 (13)
4 (7)
6 (9)
8 (9)
5 (5)
6 (9)
Absent
78 (91)
50 (91)
52 (80)
Lip
0 (0)
2 (4)
0 (0)
Present
8 (9)
5 (9)
13 (20)
0.096
*
pT1
30 (35)
7 (13)
9 (14)
pT2
43 (50)
36 (65)
39 (60)
pT3
10 (12)
7 (13)
9 (14)
pT4
3 (3)
5 (9)
8 (12)
0.014*
Lymph node metastasis Negative (NO)
74 (86)
36 (65)
36 (55)
Positive (N1–N3) Tumor stage
12 (14)
19 (35)
29 (45)
I
29 (34)
7 (13)
8 (12)
II
38 (44)
29 (53)
30 (46)
III
14 (16)
9 (16)
15 (23)
IV
5 (6)
10 (18)
12 (18)
Well
64 (74)
42 (76)
50 (77)
Moderate
14 (16)
8 (15)
12 (18)
8 (9)
5 (9)
3 (5)
1
4 (5)
0 (0)
0 (0)
2
14 (16)
5 (9)
5 (8)
3
55 (64)
30 (55)
41 (63)
4c
12 (14)
19 (35)
16 (25)
4d
1 (1)
1 (2)
3 (5)
82 (95)
51 (93)
53 (82)
4 (5)
4 (7)
12 (18)
Absent
78 (91)
46 (84)
56 (86)
Present
8 (9)
9 (16)
9 (14)
0.0001*
0.006*
Tumor differentiation 0.82
Mode of invasion
Present
0.0025*
Local recurrence
Pathologic T classification
Vascular invasion Absent
0.005*
Regional recurrence
Floor of mouth
Poor
0.33
Vital status at follow-up
Sex Female
1 (1)
p
0.028*
0.014*
Lymphatic permeation 0.44
Statistically significant (Chi-square test), single-positive group consisted of MUC1(?)/MUC4(-) and MUC1(-)/MUC4(?) patients
single-positive group (67.0 %) (p = 0.0005, p = 0.0009, respectively). There was no significant difference between the single-positive and the double-positive groups on overall survival (p = 0.80). The correlation between MUC1/MUC4 co-expression and the overall survival rate as determined by the Kaplan–Meier method is illustrated in Fig. 3. The MUC1/ MUC4 double-positive group showed the poorest outcomes especially within 60 months after surgery, whereas patients with MUC1/MUC4 double-negative expression showed the most favorable prognosis. In the univariate analysis, patients with MUC1/MUC4 single- and double-positivity showed significant risk for overall survival when compared with the double-negative group (single-positive group, HR 4.29, 95 % CI 1.71–12.13, p = 0.0016; double-positive group, HR 4.73, 95 % CI 1.96–13.12, p = 0.0004, respectively). The multivariate analysis of overall survival by a proportional hazards regression model is summarized in Table 4. Multivariate analysis of the seven prognostic factors revealed that nodal metastasis (p = 0.029), poor differentiation (p = 0.026), and MUC1/MUC4 single- and doubleexpression (p = 0.008 and 0.007, respectively) were independent prognostic factors for overall survival. Among these significant variables, the MUC1/MUC4 doublepositive group showed the strongest hazard ratio and p value (HR = 3.60, p = 0.007), demonstrating its utility as a predictor for overall survival of patients with OSCC.
123
Int J Clin Oncol
Fig. 2 Associations between MUC1/MUC4 co-expression and clinicopathological characteristics. Tumor size (a, p = 0.014), nodal metastasis (b, p = 0.0001), tumor stage (c, p = 0.006), vascular invasion (d, p = 0.014), tumor death (e, p = 0.005) and regional
recurrence (f, p = 0.0025) were significantly correlated with the MUC1/MUC4 co-expression. Tumor aggressiveness was associated with MUC1/MUC4 with an increase of positivity
Clinical significance of MUC1/MUC4 combination for disease-free survival
favorable disease-free survival compared with that of other groups, whereas MUC1/MUC4 double-positive group showed the poorest survival. Along with overall survival, in the univariate analysis, patients with MUC1/MUC4 single- and double-positivity showed a significant hazard ratio for disease-free survival (single-positive group, HR 2.70, 95 % CI 1.40–5.33, p = 0.003; double-positive group, HR 3.71, 95 % CI 2.04–7.06, p \ 0.0001, respectively). The multivariate analysis of disease-free survival is summarized in Table 5. Multivariate analysis revealed that nodal metastasis (p = 0.022), vascular invasion (p = 0.0094), and MUC1/ MUC4 single- and double-positive (p = 0.013 and p = 0.0019) were independent prognostic factors for disease-free
The 5-year disease-free survival rate of all patients was 63.7 %, and that of the MUC1/MUC4 double-negative group was high (81.1 %) and significantly better than that of the double-positive group (45.7 %) and single-positive group (56.1 %) (p \ 0.0001, p = 0.002, respectively). There was no significant difference between the singlepositive and the double-positive groups on disease-free survival (p = 0.24). The correlation between MUC1/ MUC4 expression and disease-free survival rate as determined by the Kaplan–Meier method is illustrated in Fig. 4. The MUC1/MUC4 double-negative group showed obvious
123
Int J Clin Oncol
Fig. 3 Correlation between MUC1/MUC4 co-expression and overall survival. The overall survival of patients with double-negative expression of MUC1/MUC4 was clearly better than that of the other patient groups composed of single-positive and double-positive group (p = 0.001). Within 60 months after surgery, the double-positive group showed worse survival than the single-positive group
Fig. 4 Correlation between MUC1/MUC4 co-expression and disease-free survival. The disease-free survival of patients with doublenegative expression of MUC1/MUC4 was statistically better than that of other groups composed of single-positive and double-positive group (p \ 0.0001), whereas MUC1/MUC4 double-positive group showed the poorest disease-free survival
Table 4 Multivariate analysis of overall survival
Table 5 Multivariate analysis of disease-free survival
Variable
HR
95 % CI
p
Tumor location
HR
95 % CI
p
1.10–3.26
0.022*
0.78–2.21
0.28
1.26–4.41
0.0094*
Lymph node metastasis
Tongue, mandible, buccal, and lip
1
Maxilla and floor of mouth
1.68
0.84–3.35
0.13
Pathologic T classification
Negative (N0)
1
Positive (N1–N3)
1.93
Mode of invasion
T1 or T2
1
T3 or T4
1.53
0.71–3.08
0.27
Lymph node metastasis
1–3
1
Diffuse invasive (4c or 4d)
1.33
Vascular invasion
Negative (N0)
1
Positive (N1–N3)
2.24
1.09–4.43
0.029*
Tumor differentiation Well or moderate Poor
Variable
Absent
1
Present
2.43
MUC1/MUC4 expression 1 3.02
1.15–7.06
0.026*
Mode of invasion 1–3
1
Diffuse invasive (4c or 4d)
1.57
Double-negative (-/-) Single-positive (?/- or -/?)
1 2.33
1.20–4.66
0.013*
Double-positive (?/?)
2.71
1.44–5.32
0.0019*
HR hazard ratio, 95 % CI 95 % confidence interval 0.77–3.14
0.21
*
0.92–5.13
0.074
Analysis of risk factors correlated with regional recurrence
Statistically significant
Vascular invasion Absent
1
Present
2.27
MUC1/MUC4 expression Double-negative (-/-)
1
Single-positive (?/- or -/?)
3.49
1.37–9.98
0.008*
Double-positive (?/?)
3.60
1.40–10.42
0.007*
HR hazard ratio, 95 % CI 95 % confidence interval *
Statistically significant
survival. Taken together, double-positivity/negativity of MUC1/MUC4 expression was a powerful predictor for poor/ favorable disease-free survival of patients with OSCC.
Regional recurrence occurred with the increase of MUC1/ MUC4 positivity; this occurred in six of the 86 (7 %) patients in MUC1/MUC4 double-negative group, in 12 of 55 (22 %) patients in single-positive group, and in 18 of 65 (28 %) patients in double-positive group (p = 0.0003, Chisquare test, Table 3). Multivariate analysis of variables correlated with regional recurrence is summarized in Table 6. In multivariate analysis, MUC1/MUC4 coexpression was found to be an independent risk factor (single-positive group, HR 3.29, 95 % CI 1.16–10.19,
123
Int J Clin Oncol Table 6 Multivariate analysis of variables correlated with subsequent lymph node metastasis Variable
HR
95 % CI
p
0.73–3.69
0.22
0.65–5.55
0.23
Mode of invasion 1–3
1
Diffuse invasive (4c or 4d)
1.66
Vascular invasion Absent
1
Present
1.96
MUC1/MUC4 expression Double-negative (-/-)
1
Single-positive (?/- or -/?)
3.29
1.16–10.19
0.025*
Double-positive (?/?)
4.35
1.65–12.90
0.0025*
HR hazard ratio, 95 % CI 95 % confidence interval *
Statistically significant
p = 0.025; double-positive group, HR 4.35, 95 % CI 1.65–12.90, p = 0.0025, respectively). Among all factors evaluated in this study, MUC1/MUC4 double-positivity was considered to be a most strong marker to predict regional recurrence in OSCC patients.
Discussion Currently, the indications for elective neck dissection in clinically N0 patients with OSCC remain controversial. Regional recurrence is one of the main causes of death from OSCC [4, 5]. However, neck dissection often results in serious complications. In this study, our results showed that double-negativity for MUC1/MUC4 strongly correlated with better outcomes after surgery and that MUC1/ MUC4 double-positivity was the strong independent prognostic factor for regional recurrence. Therefore, the co-expression profile of MUC1/MUC4 may be a potent indication for elective neck dissection. Several previous studies have demonstrated the usefulness of MUC1 and MUC4 in cancer detection and prognosis. Implication of mucins in the pathogenesis as oncogenes has provided them as highly attractive targets for cancer therapy [9]. Measurement of circulating MUC1N levels has been used to monitor the clinical course of patients with breast cancer and to detect early disease recurrence [9, 35]. On the other hand, MUC4 was found to mask the antibody-binding epitope of ErbB2 and lead to diminished Herceptin binding, suggesting an interesting attribute of MUC4 that may prove useful in therapeutic planning [36, 37]. The present study demonstrated that MUC1/MUC4 combination was a powerful indicator for tumor aggressiveness and poor outcome in OSCC. The MUC1/MUC4 double-positive group showed the poorest
123
overall and disease-free survival, whereas patients with MUC1/MUC4 double-negative expression showed the most favorable prognosis. Multivariate analysis revealed that MUC1/MUC4 double-positivity was a strong independent prognostic factor for overall and disease-free survival. This is the first report to provide evidence of the clinical prognostic value of the combination of MUC1 and MUC4 expression in OSCC and suggests that determination of the combined MUC1/MUC4 expression status in biopsy specimens has the potential to influence decisions regarding treatment. Furthermore, these mucins may be attractive targets for the therapy in OSCC in the future. Similar to this study, previous immunohistochemical studies of intrahepatic cholangiocarcinoma-mass forming type [10] and gallbladder carcinoma [32] have attempted to reveal combinations of mucins as prognostic indicators. Shibahara et al. [10] have demonstrated the tendency of MUC1/MUC4 co-expression (p = 0.065) and that patients with MUC1/MUC4 double-positive expression show the poorest outcomes (100 % death within 20 months after surgery), whereas patients with MUC1/MUC4 doublenegative expression show better outcomes (p = 0.0004) in intrahepatic cholangiocarcinoma-mass forming type. Furthermore, other immunohistochemical studies using MUC1, MUC2 and MUC4 in gallbladder carcinoma showed a significant correlation between MUC4 and MUC1 expression (p = 0.004), and that patients with a MUC4(?)/MUC2(-) pattern show the poorest outcomes compared with those of patients with other combinations of MUC4/MUC2 expression (p \ 0.05) [32]. In this manner, the combination of several mucin-expression profiles may be used to represent the character of tumors for patient prognosis because mucin expression varies considerably depending on the tissues, organs, and tumors [23]. The biological mechanism of co-expression of MUC1 and MUC4 is unclear at present. To date, epigenetic regulations such as DNA methylation and histone modification have been reported to affect both MUC1 [38] and MUC4 [39, 40] expression, suggesting that the regulatory mechanism of this co-expression may be caused by aberrant status of epigenome. Therefore, comprehensive analysis including mRNA and protein expression and epigenetic aberrations of MUC1 and MUC4 in OSCC would be an interesting area of future study. Moreover, to date, many variants of alternative splicing of the MUC1 transcript have also been observed [41] and some MUC1 splice variants are expressed in human neoplasms such as ovarian tumors [42] and ductal carcinoma in situ of the breast [43]. Interestingly, Obermair et al. [42] have demonstrated that specific MUC1 splice variants are associated with tumor malignancy, suggesting its usefulness as a diagnostic tool. Also, alternative splicing of the MUC4 has been reported [44] and some of these are expressed in pancreatic
Int J Clin Oncol
carcinomas but not in the normal pancreas [45, 46]. Taken together, the dysfunction of these regulations of MUC1 and MUC4 expression such as epigenetic alterations and alternative splicing might be a candidate for a molecular marker and a therapeutic target of OSCC in the future. In conclusion, this study demonstrated significant coexpression of MUC1/MUC4 and that the clinical prognostic value of the combination of MUC1 and MUC4 expression in OSCC. The results suggest that determination of the co-expression status of MUC1/MUC4 in biopsy specimens has the potential to influence decisions regarding treatment, such as the need for elective neck dissection. Acknowledgments The authors thank Mr. Yoshiharu Atsuji and Ms. Yukari Nishimura for their excellent technical assistance. This study was supported in part by Grants-in-Aid for Scientific Research (B) 23390085 to S. Yonezawa; Scientific Research (C) 24590447 to M. Higashi; Scientific Research (B) 23390466 to K. Sugihara and Young Scientists (B) 24792239 to T. Hamada from the Ministry of Education, Science, Sports, Culture and Technology, Japan.
13.
14.
15.
16.
17.
18.
19. Conflict of interest The authors declare that they have no conflicts of interest. 20.
21.
References 1. Jemal A, Bray F, Center MM et al (2011) Global cancer statistics. CA Cancer J Clin 61:69–90 2. Warnakulasuriya S (2009) Global epidemiology of oral and oropharyngeal cancer. Oral Oncol 45:309–316 3. Cheng A, Schmidt BL (2008) Management of the N0 neck in oral squamous cell carcinoma. Oral Maxillofac Surg Clin N Am 20:477–497 4. Kowalski LP, Bagietto R, Lara JR et al (2000) Prognostic significance of the distribution of neck node metastasis from oral carcinoma. Head Neck 22:207–214 5. Capote A, Escorial V, Munoz-Guerra MF et al (2007) Elective neck dissection in early-stage oral squamous cell carcinoma— does it influence recurrence and survival? Head Neck 29:3–11 6. Mahfouz ME, Rodrigo JP, Takes RP et al (2010) Current potential and limitations of molecular diagnostic methods in head and neck cancer. Eur Arch Otorhinolaryngol 267:851–860 7. Hollingsworth MA, Swanson BJ (2004) Mucins in cancer: protection and control of the cell surface. Nat Rev Cancer 4:45–60 8. Moniaux N, Escande F, Porchet N et al (2001) Structural organization and classification of the human mucin genes. Front Biosci 6:1192–1206 9. Kufe DW (2009) Mucins in cancer: function, prognosis and therapy. Nat Rev Cancer 9:874–885 10. Shibahara H, Tamada S, Higashi M et al (2004) MUC4 is a novel prognostic factor of intrahepatic cholangiocarcinoma-mass forming type. Hepatology 39:220–229 11. Higashi M, Yonezawa S, Ho JJ et al (1999) Expression of MUC1 and MUC2 mucin antigens in intrahepatic bile duct tumors: its relationship with a new morphological classification of cholangiocarcinoma. Hepatology 30:1347–1355 12. Nakamura A, Horinouchi M, Goto M et al (2002) New classification of pancreatic intraductal papillary-mucinous tumour by
22.
23.
24.
25. 26.
27.
28.
29.
30.
31.
mucin expression: its relationship with potential for malignancy. J Pathol 197:201–210 Kitazono I, Higashi M, Kitamoto S et al (2013) Expression of MUC4 mucin is observed mainly in the intestinal type of intraductal papillary mucinous neoplasm of the pancreas. Pancreas 42:1120–1128 Osako M, Yonezawa S, Siddiki B et al (1993) Immunohistochemical study of mucin carbohydrates and core proteins in human pancreatic tumors. Cancer 71:2191–2199 Sagara M, Yonezawa S, Nagata K et al (1999) Expression of mucin 1 (MUC1) in esophageal squamous-cell carcinoma: its relationship with prognosis. Int J Cancer 84:251–257 Takao S, Uchikura K, Yonezawa S et al (1999) Mucin core protein expression in extrahepatic bile duct carcinoma is associated with metastases to the liver and poor prognosis. Cancer 86:1966–1975 Saitou M, Goto M, Horinouchi M et al (2005) MUC4 expression is a novel prognostic factor in patients with invasive ductal carcinoma of the pancreas. J Clin Pathol 58:845–852 Shibahara H, Tamada S, Goto M et al (2004) Pathologic features of mucin-producing bile duct tumors: two histopathologic categories as counterparts of pancreatic intraductal papillary-mucinous neoplasms. Am J Surg Pathol 28:327–338 Tamada S, Goto M, Nomoto M et al (2002) Expression of MUC1 and MUC2 mucins in extrahepatic bile duct carcinomas: its relationship with tumor progression and prognosis. Pathol Int 52:713–723 Tamada S, Shibahara H, Higashi M et al (2006) MUC4 is a novel prognostic factor of extrahepatic bile duct carcinoma. Clin Cancer Res 12:4257–4264 Tamura Y, Higashi M, Kitamoto S et al (2012) MUC4 and MUC1 expression in adenocarcinoma of the stomach correlates with vessel invasion and lymph node metastasis: an immunohistochemical study of early gastric cancer. PLoS One 7:e49251 Tsutsumida H, Goto M, Kitajima S et al (2007) MUC4 expression correlates with poor prognosis in small-sized lung adenocarcinoma. Lung Cancer 55:195–203 Yonezawa S, Goto M, Yamada N et al (2008) Expression profiles of MUC1, MUC2, and MUC4 mucins in human neoplasms and their relationship with biological behavior. Proteomics 8:3329–3341 Yonezawa S, Higashi M, Yamada N et al (2010) Significance of mucin expression in pancreatobiliary neoplasms. J Hepatobiliary Pancreat Sci 17:108–124 Yonezawa S, Higashi M, Yamada N et al (2008) Precursor lesions of pancreatic cancer. Gut Liver 2:137–154 Yonezawa S, Taira M, Osako M et al (1998) MUC-1 mucin expression in invasive areas of intraductal papillary mucinous tumors of the pancreas. Pathol Int 48:319–322 Yonezawa S, Higashi M, Yamada N et al (2011) Mucins in human neoplasms: clinical pathology, gene expression and diagnostic application. Pathol Int 61:697–716 Yonezawa S, Sato E (1997) Expression of mucin antigens in human cancers and its relationship with malignancy potential. Pathol Int 47:813–830 Yonezawa S, Nakamura A, Horinouchi M et al (2002) The expression of several types of mucin is related to the biological behavior of pancreatic neoplasms. J Hepatobiliary Pancreat Surg 9:328–341 Hamada T, Nomura M, Kamikawa Y et al (2012) DF3 epitope expression on MUC1 mucin is associated with tumor aggressiveness, subsequent lymph node metastasis, and poor prognosis in patients with oral squamous cell carcinoma. Cancer 118:5251–5264 Hamada T, Wakamatsu T, Miyahara M et al (2012) MUC4: a novel prognostic factor of oral squamous cell carcinoma. Int J Cancer 130:1768–1776
123
Int J Clin Oncol 32. Lee HK, Cho MS, Kim TH (2012) Prognostic significance of muc4 expression in gallbladder carcinoma. World J Surg Oncol 10:224 33. Sobin LGM, Wittekind Ce (2009) TNM Classification of malignant tumors. In: Hoboken NJ (ed), 7th ed John Wiley & Sons, Inc. 34. Yamamoto E, Kohama G, Sunakawa H et al (1983) Mode of invasion, bleomycin sensitivity, and clinical course in squamous cell carcinoma of the oral cavity. Cancer 51:2175–2180 35. Hayes DF, Sekine H, Ohno T et al (1985) Use of a murine monoclonal antibody for detection of circulating plasma DF3 antigen levels in breast cancer patients. J Clin Invest 75:1671–1678 36. Nagy P, Friedla¨nder E, Tanner M et al (2005) Decreased accessibility and lack of activation of ErbB2 in JIMT-1, a Herceptinresistant, MUC4-expressing breast cancer cell line. Cancer Res 65:473–482 37. Singh AP, Chaturvedi P, Batra SK (2007) Emerging roles of MUC4 in cancer: a novel target for diagnosis and therapy. Cancer Res 67:433–436 38. Yamada N, Nishida Y, Tsutsumida H et al (2008) MUC1 expression is regulated by DNA methylation and histone H3 lysine 9 modification in cancer cells. Cancer Res 68:2708–2716 39. Yamada N, Nishida Y, Tsutsumida H et al (2009) Promoter CpG methylation in cancer cells contributes to the regulation of MUC4. Br J Cancer 100:344–351
123
40. Vincent A, Ducourouble MP, Van Seuningen I (2008) Epigenetic regulation of the human mucin gene MUC4 in epithelial cancer cell lines involves both DNA methylation and histone modifications mediated by DNA methyltransferases and histone deacetylases. FASEB J 22:3035–3045 41. Zhang L, Vlad A, Milcarek C et al (2013) Human mucin MUC1 RNA undergoes different types of alternative splicing resulting in multiple isoforms. Cancer Immunol Immunother 62:423–435 42. Obermair A, Schmid BC, Packer LM et al (2002) Expression of MUC1 splice variants in benign and malignant ovarian tumours. Int J Cancer 100:166–171 43. Schmid BC, Rudas M, Fabjani G et al (2003) Evaluation of MUC1 splice variants as prognostic markers in patients with ductal carcinoma in situ of the breast. Oncol Rep 10:1981–1985 44. Choudhury A, Moniaux N, Winpenny JP et al (2000) Human MUC4 mucin cDNA and its variants in pancreatic carcinoma. J Biochem 128:233–243 45. Moniaux N, Escande F, Batra SK et al (2000) Alternative splicing generates a family of putative secreted and membrane-associated MUC4 mucins. Eur J Biochem 267:4536–4544 46. Choudhury A, Moniaux N, Ringel J et al (2001) Alternate splicing at the 30-end of the human pancreatic tumor-associated mucin MUC4 cDNA. Teratog Carcinog Mutagen 21:83–96
ORIGINAL ARTICLE
Combination of MUC1 and MUC4 expression predicts clinical outcome in patients with oral squamous cell carcinoma Yoshiaki Kamikawa • Yuji Kanmura • Tomofumi Hamada • Norishige Yamada • Muzafar A. Macha • Surinder K. Batra • Michiyo Higashi • Suguru Yonezawa • Kazumasa Sugihara
Received: 19 February 2014 / Accepted: 12 May 2014 Ó Japan Society of Clinical Oncology 2014
Abstract Background Both MUC1 and MUC4 are high molecular weight glycoproteins and are independent indicators of worse prognosis in many human epithelial cancers including oral squamous cell carcinoma (OSCC). However, there has been no investigation of the clinical importance of the co-expression of MUC1 and MUC4 in OSCC. The aim of this study was to evaluate the co-expression profile of MUC1/MUC4 and analyze the prognostic significance in OSCC. Methods We examined the expression profile of MUC1 and MUC4 in OSCC tissues from 206 patients using immunohistochemistry. The co-expression profile of MUC1/MUC4 and its prognostic significance in OSCC was statistically analyzed.
Y. Kamikawa and Y. Kanmura contributed equally to this work.
Results MUC1 and MUC4 overexpression were strongly correlated with each other (p \ 0.0001) and a combination of both MUC1 and MUC4 expression was a powerful indicator for tumor aggressiveness such as tumor size (p = 0.014), lymph node metastasis (0.0001), tumor stage (p = 0.006), diffuse invasion (p = 0.028), and vascular invasion (p = 0.014). The MUC1/MUC4 double-positive patients showed the poorest overall and disease-free survival. Multivariate analysis revealed that MUC1/MUC4 double-positivity was the strong independent prognostic factor for overall and disease-free survival (p = 0.007 and (p = 0.0019), in addition to regional recurrence (p = 0.0025). Conclusions Taken together, these observations indicate that the use of a combination of MUC1/MUC4 can predict outcomes for patients with OSCC. This combination is also a useful marker for predicting regional recurrence. MUC1 and MUC4 may be attractive targets for the selection of treatment methods in OSCC.
Electronic supplementary material The online version of this article (doi:10.1007/s10147-014-0710-6) contains supplementary material, which is available to authorized users. Y. Kamikawa T. Hamada (&) K. Sugihara Department of Oral Surgery, Kagoshima University Hospital, 8-35-1 Sakuragaoka, Kagoshima 890-8544, Japan e-mail: [email protected]
M. A. Macha S. K. Batra Department of Biochemistry and Molecular Biology, Buffett Cancer Center, Eppley Cancer Institute, University of Nebraska Medical Center, 985870 Nebraska Medical Center, Omaha, NE 68198-5870, USA
Y. Kanmura K. Sugihara Department of Maxillofacial Diagnostic and Surgical Science, Field of Oral and Maxillofacial Rehabilitation, Kagoshima University Graduate School of Medical and Dental Sciences, 8-35-1 Sakuragaoka, Kagoshima 890-8544, Japan
M. Higashi S. Yonezawa Department of Human Pathology, Field of Oncology, Kagoshima University Graduate School of Medical and Dental Sciences, 8-35-1 Sakuragaoka, Kagoshima 890-8544, Japan
N. Yamada Division of Reproductive Sciences, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave., Cincinnati, Ohio 45229, USA
123
Int J Clin Oncol
Keywords Oral squamous cell carcinoma MUC1 MUC4 Immunohistochemistry Prognosis
Materials and methods Patients
Introduction Oral cancer is a serious problem in many parts of the world, with approximately 263,900 new cases and 128,000 deaths worldwide in 2008, and oral squamous cell carcinomas (OSCCs) comprise [90 % of these cases [1]. Survival rates of OSCC have not improved significantly over the past 30 years, despite outstanding progress in clinical and basic research [2]. One of the main prognosticators of the survival of patients with OSCC is regional lymph node metastasis, but such metastasis is quite often subclinical or occult at the time of diagnosis [3, 4]. Additionally, patients treated with curative surgery often have poor outcomes because of unexpected events such as regional recurrence [3, 5]. Therefore prognostic factors that predict clinical outcomes may be helpful in guiding treatment decisions. However, despite the substantial research effort made over decades, very few prognostic markers and predictive assays have been established in routine clinical diagnosis and treatment of OSCC [6]. Mucins are high molecular weight glycoproteins that help to protect and lubricate luminal surfaces under normal physiological conditions [7, 8]. To date, 21 different mucins have been found in humans and functionally classified into secreted and membrane-bound types. These mucins are involved in cell differentiation, renewal of the epithelium and modulation of cell adhesion, immune response, and cell signaling [7–9]. Our previous studies have reported the deregulation of both MUC1 and MUC4 in several human malignancies and these are associated with aggressive tumor behavior and poor outcomes [10–29]. In OSCC, aberrant expression of MUC1 [30] or MUC4 [31] is an independent prognostic factor indicating a poor prognosis. Interestingly, previous studies of intrahepatic cholangiocarcinoma-mass forming type [10] and gallbladder carcinoma [32] demonstrated a significant correlation between MUC1 and MUC4 expression, although these mucins are encoded by different genes (MUC1, chromosome 1q21; MUC4, chromosome 3q29) and are involved in different signaling pathways [7– 9]. However, co-expression status and its clinical significance of MUC1 and MUC4 have not been investigated in OSCCs. Therefore, in the present study, we examined the expression profile of MUC1 and MUC4 and found their coexpression in OSCC tissues to be significant. Furthermore, MUC1/MUC4 co-expression was correlated with tumor aggressiveness, and MUC1/MUC4 double-positivity was observed to be an independent worse prognostic factor with decreased overall and disease-free survival, in addition to regional recurrence in patients with OSCC.
123
In the present study, we used the patient series collected for our previous studies of MUC1 [30] and MUC4 [31] expression in OSCC. The information of vital status of all patients was updated and survival rates were re-calculated. This study was conducted in accordance with the guiding principles of the Declaration of Helsinki. After ethics approval from the Human Ethics Committee of Kagoshima University Graduate School of Medical and Dental Sciences (permission no. 34, 65, 99, 101), a retrospective search was carried out to identify patients pathologically diagnosed with OSCC. This study included 206 patients who underwent surgical resection for primary oral cancer between 1992 and 2008 at the Department of Oral Surgery, Kagoshima University Hospital in Kagoshima, Japan. Seventy-eight out of 206 (38 %) patients received neoadjuvant chemotherapy or postoperative chemotherapy that used fluorouracil or platinum-containing anti-cancer drugs such as CDDP and CBDCA. There were 128 out of 206 (62 %) patients that were initially treated with only surgical resection. Patients were excluded if they had received previous treatment for oral cancer, had distant metastasis at the time of their first visit, or did not undergo surgical treatment. Informed consent was obtained to use surgical specimens for this study from all of the patients. Histopathological analysis All specimens were fixed in formalin, embedded in paraffin, and cut into 4-lm-thick sections for hematoxylin and eosin (HE) staining for histopathological analysis and for immunohistochemical analysis of MUC1 and MUC4. Detailed tumor characteristics, including the tumor-nodemetastasis (TNM) classification, differentiation, mode of invasion, vascular invasion, lymphatic permeation, and perineural infiltration were evaluated histologically. Anatomical sites, tumor differentiation, TNM classification, and stage grouping of the tumor were classified according to the TNM classification of malignant tumors [33]. Aggressiveness of invasion was assessed by classification of the mode of invasion [34]. Immunohistochemistry Immunohistochemical staining was performed as described previously [30, 31]. Briefly, for MUC1 staining, we used a mouse monoclonal antibody DF3 (Toray-Fuji Bionics, Tokyo, Japan) and MUC4 was detected by a mouse monoclonal antibody8G7 (generated by one of us, S. K. Batra). Antigen retrieval was performed using CC1
Int J Clin Oncol
antigen retrieval buffer (Ventana Medical Systems, Tucson, AZ, USA) for MUC1 and by 0.01 M citrate buffer (pH 6.0) in water bath at 80 °C for 20 min for MUC4. After washing with PBS, tissue sections were incubated at 37 °C for 32 min for DF3 (diluted at 1:10) [30] and at 4 °C for 16 h for 8G7 (diluted at 1:3,000) [31]. Primary antibody was replaced by hybridoma culture medium, normal mouse serum, or PBS in negative controls. Two blinded investigators (Y. K. and T. H.) independently evaluated the results of immunostaining without knowledge of the patient outcomes. According to preliminary statistical analysis by receiver operating characteristic (ROC) curve, specimens showing more than 5 % positively stained carcinoma cells were graded as positive [30, 31]. Statistical analysis Associations between mucin expression and clinicopathological characteristics were evaluated by the Chi-square test. The accuracy of the combination of MUC1/MUC4 positivity was evaluated from the ROC curve analysis by calculating the area under the curve (AUC). A ROC curve was constructed by plotting the sensitivity (true-positive rate) against 1-specificity (false-positive rate) for detection of tumor death and recurrence. Overall and disease-free survival of the patients was compared between groups according to the Kaplan–Meier method. Differences in survival rates were evaluated using the log-rank test. Overall survival was measured from the date of surgery to the date of the last follow-up examination or death, whereas disease-free survival was measured from the date of surgery to the date that local recurrence and/or regional recurrence were first detected. We defined ‘‘regional recurrence’’ as regional lymph node metastasis that is detected after initial treatment in all patients regardless of neck dissection. For analysis of prognostic factors, six tumor locations were classified into one of two groups: the better survival group (lip, buccal, tongue, and mandible, n = 157) and the poor survival group (maxilla and floor of the mouth, n = 49) according to preliminary survival analyses described previously [31]. Univariate and multivariate analyses of overall and disease-free survival were performed using the Cox proportional hazards regression model. Variables that showed p \ 0.10 in the univariate analysis in a previous study [30] were used to create a multivariate model in this study to determine whether mucin expression was an independent prognostic factor. In multivariate analysis, the tumor stage was excluded for the model, because the tumor stage contains both T and N classifications, which would negate the statistical significance of each other. In addition to survival analysis, we performed univariate and multivariate analyses of the relative risk for regional recurrence.
Statistical analysis was performed using JMP software (version 9.0.2; SAS Institute, Cary, NC, USA). All statistics were two-sided, and a probability of p \ 0.05 was considered to be significant.
Results MUC1 and MUC4 expression and their correlation in OSCC Expression of MUC1 and MUC4 was analyzed in OSCC tissues by IHC and representative patterns of combination of expression with HE staining as reference are illustrated in Fig. 1. The statistical correlation between MUC1 and MUC4 expressions in 206 OSCC patients is shown in Table 1. There were 95/206 (46 %) MUC1-positive and 80/206 (39 %) MUC4-positive patients. Double-positivity of MUC1 and MUC4 was observed in 60 of 206 (29 %) OSCC cases (Fig. 1a), whereas double negativity was found in 91 (44 %) patients (Fig. 1d). Taken together, MUC1 and MUC4 expression was strongly correlated with each other (p \ 0.0001, Table 1). Combinations of MUC1 and MUC4 expression in OSCC To determine the most appropriate combination of MUC1 and MUC4 expression as a prognostic factor in OSCC, we performed ROC analysis and calculated the AUC (Table 2). When patients were divided into three groups (MUC1/MUC4 double-positive (?/?) group (Fig. 1a), single-positive (?/- or -/?) group (Fig. 1b, c), and double-negative (-/-) group (Fig. 1d), the AUC for tumor death was 0.658 and that for recurrence was 0.668. These AUCs were higher than those of MUC1 or MUC4 single analysis (0.625 and 0.616 for MUC1, 0.614 and 0.639 for MUC4, respectively), suggesting that the combination of MUC1 and MUC4 is a more useful prognostic factor than single evaluation to predict the outcomes of patients with OSCC. Therefore, we used ‘‘MUC1/MUC4 double-positive vs. single-positive vs. double-negative’’ as the grouping method for further analysis. Association between MUC1/MUC4 co-expression and clinicopathologic characteristics Associations between MUC1/MUC4 co-expression and clinicopathological characteristics are summarized in Table 3, and factors with statistical significance are illustrated in Fig. 2. The percentages of patients with a large tumor size (p = 0.014), positive nodal metastasis (p = 0.0001), advanced tumor stage (p = 0.006), diffuse invasion
123
Int J Clin Oncol Fig. 1 Representative patterns of combination of MUC1/ MUC4 immunohistochemistry. a Double-positive group (MUC1?/MUC4?, n = 60 of total 206 patients). Both MUC1 and MUC4 showed strong positivity in invasive carcinoma cells just beneath epithelium. MUC4 was focally positive in dysplastic epithelium in this case. b Single-positive group (MUC1?/MUC4-, n = 35). Tumor cells showed MUC1positive but MUC4-negative expression in this case. c Singlepositive group (MUC1-/ MUC4?, n = 20). c Contrary to b, tumor cells showed MUC1negative but MUC4-positive expression in this case. d Double negative group (MUC1-/MUC4-, n = 91). The tumor cells were negative for both MUC1 and MUC4
Table 1 Correlation (n = 206) MUC1 expression
Negative
between
MUC1
MUC4 expression Negative
Positive
91
20
and
MUC4
expression
Total
Table 2 The accuracy from combination of MUC1 and MUC4 Grouping method
Tumor death
Local or neck recurrence
0.625
0.616
0.614
0.639
Double-positive vs. other
0.586
0.612
Double-negative vs. other
0.654
0.643
Double-negative vs. single-positive vs. double-positive
0.658
0.668
111
Positive
35
60
95
Total
126
80
206 p \ 0.0001 (Chi-square)
MUC1 Positive vs. negative MUC4 Positive vs. negative
of cancer cells (p = 0.028), vascular invasion (p = 0.014), tumor death (p = 0.005), and regional recurrence (p = 0.0025) were significantly correlated with MUC1/ MUC4 co-expression. As shown in Fig. 2, tumor aggressiveness was strongly associated with MUC1/MUC4 double-positivity rather than single-positivity. Overall, the use of MUC1/MUC4 combination was a powerful indicator for tumor aggressiveness and poor outcome in OSCC. There was no significant difference in the proportion of patients who received chemotherapy between the MUC1-positive group and the MUC1-negative group and between the MUC4-positive group and the MUC4-negative group. Also, there was no significant difference in the type of anti-cancer drug between these groups.
123
AUC in ROC analysis endpoint
MUC1/MUC4
AUC area under the curve, ROC receiver operating characteristic
Clinical significance of MUC1/MUC4 combination for overall survival The 5-year overall survival rate of all patients was 79.4 %. The 5-year overall survival rate of the MUC1/MUC4 doublenegative group was obviously high (92.2 %) and significantly better than that of the double-positive group (72.0 %) and the
Int J Clin Oncol Table 3 Association between MUC1/MUC4 co-expression and clinicopathologic factors Variable
MUC1/MUC4 expression no. of patients (%) Doublenegative n = 86
Singlepositive n = 55
Table 3 continued Variable
p
MUC1/MUC4 expression no. of patients (%) Doublenegative n = 86
Singlepositive n = 55
Doublepositive n = 65
52 (95)
63 (97)
3 (5)
2 (3)
75 (87)
39 (71)
45 (69)
Death from OSCC
6 (7)
13 (24)
19 (29)
Death from other than OSCC
5 (6)
3 (5)
1 (2)
Doublepositive n = 65
Perineural infiltration Absent 85 (99)
Age (years) \65
38 (44)
28 (51)
25 (38)
C65
48 (56)
27 (49)
40 (62)
30 (35)
17 (31)
26 (40)
56 (65)
38 (69)
39 (60)
Tongue
35 (41)
26 (47)
31 (48)
Mandible
16 (19)
13 (24)
13 (20)
Absent
80 (93)
43 (78)
47 (72)
Maxilla
6 (19)
5 (9)
9 (14)
Present
6 (7)
12 (22)
18 (28)
Buccal
0.39
Present
Male Tumor location
Alive 0.58
0.47
11 (13)
4 (7)
6 (9)
8 (9)
5 (5)
6 (9)
Absent
78 (91)
50 (91)
52 (80)
Lip
0 (0)
2 (4)
0 (0)
Present
8 (9)
5 (9)
13 (20)
0.096
*
pT1
30 (35)
7 (13)
9 (14)
pT2
43 (50)
36 (65)
39 (60)
pT3
10 (12)
7 (13)
9 (14)
pT4
3 (3)
5 (9)
8 (12)
0.014*
Lymph node metastasis Negative (NO)
74 (86)
36 (65)
36 (55)
Positive (N1–N3) Tumor stage
12 (14)
19 (35)
29 (45)
I
29 (34)
7 (13)
8 (12)
II
38 (44)
29 (53)
30 (46)
III
14 (16)
9 (16)
15 (23)
IV
5 (6)
10 (18)
12 (18)
Well
64 (74)
42 (76)
50 (77)
Moderate
14 (16)
8 (15)
12 (18)
8 (9)
5 (9)
3 (5)
1
4 (5)
0 (0)
0 (0)
2
14 (16)
5 (9)
5 (8)
3
55 (64)
30 (55)
41 (63)
4c
12 (14)
19 (35)
16 (25)
4d
1 (1)
1 (2)
3 (5)
82 (95)
51 (93)
53 (82)
4 (5)
4 (7)
12 (18)
Absent
78 (91)
46 (84)
56 (86)
Present
8 (9)
9 (16)
9 (14)
0.0001*
0.006*
Tumor differentiation 0.82
Mode of invasion
Present
0.0025*
Local recurrence
Pathologic T classification
Vascular invasion Absent
0.005*
Regional recurrence
Floor of mouth
Poor
0.33
Vital status at follow-up
Sex Female
1 (1)
p
0.028*
0.014*
Lymphatic permeation 0.44
Statistically significant (Chi-square test), single-positive group consisted of MUC1(?)/MUC4(-) and MUC1(-)/MUC4(?) patients
single-positive group (67.0 %) (p = 0.0005, p = 0.0009, respectively). There was no significant difference between the single-positive and the double-positive groups on overall survival (p = 0.80). The correlation between MUC1/MUC4 co-expression and the overall survival rate as determined by the Kaplan–Meier method is illustrated in Fig. 3. The MUC1/ MUC4 double-positive group showed the poorest outcomes especially within 60 months after surgery, whereas patients with MUC1/MUC4 double-negative expression showed the most favorable prognosis. In the univariate analysis, patients with MUC1/MUC4 single- and double-positivity showed significant risk for overall survival when compared with the double-negative group (single-positive group, HR 4.29, 95 % CI 1.71–12.13, p = 0.0016; double-positive group, HR 4.73, 95 % CI 1.96–13.12, p = 0.0004, respectively). The multivariate analysis of overall survival by a proportional hazards regression model is summarized in Table 4. Multivariate analysis of the seven prognostic factors revealed that nodal metastasis (p = 0.029), poor differentiation (p = 0.026), and MUC1/MUC4 single- and doubleexpression (p = 0.008 and 0.007, respectively) were independent prognostic factors for overall survival. Among these significant variables, the MUC1/MUC4 doublepositive group showed the strongest hazard ratio and p value (HR = 3.60, p = 0.007), demonstrating its utility as a predictor for overall survival of patients with OSCC.
123
Int J Clin Oncol
Fig. 2 Associations between MUC1/MUC4 co-expression and clinicopathological characteristics. Tumor size (a, p = 0.014), nodal metastasis (b, p = 0.0001), tumor stage (c, p = 0.006), vascular invasion (d, p = 0.014), tumor death (e, p = 0.005) and regional
recurrence (f, p = 0.0025) were significantly correlated with the MUC1/MUC4 co-expression. Tumor aggressiveness was associated with MUC1/MUC4 with an increase of positivity
Clinical significance of MUC1/MUC4 combination for disease-free survival
favorable disease-free survival compared with that of other groups, whereas MUC1/MUC4 double-positive group showed the poorest survival. Along with overall survival, in the univariate analysis, patients with MUC1/MUC4 single- and double-positivity showed a significant hazard ratio for disease-free survival (single-positive group, HR 2.70, 95 % CI 1.40–5.33, p = 0.003; double-positive group, HR 3.71, 95 % CI 2.04–7.06, p \ 0.0001, respectively). The multivariate analysis of disease-free survival is summarized in Table 5. Multivariate analysis revealed that nodal metastasis (p = 0.022), vascular invasion (p = 0.0094), and MUC1/ MUC4 single- and double-positive (p = 0.013 and p = 0.0019) were independent prognostic factors for disease-free
The 5-year disease-free survival rate of all patients was 63.7 %, and that of the MUC1/MUC4 double-negative group was high (81.1 %) and significantly better than that of the double-positive group (45.7 %) and single-positive group (56.1 %) (p \ 0.0001, p = 0.002, respectively). There was no significant difference between the singlepositive and the double-positive groups on disease-free survival (p = 0.24). The correlation between MUC1/ MUC4 expression and disease-free survival rate as determined by the Kaplan–Meier method is illustrated in Fig. 4. The MUC1/MUC4 double-negative group showed obvious
123
Int J Clin Oncol
Fig. 3 Correlation between MUC1/MUC4 co-expression and overall survival. The overall survival of patients with double-negative expression of MUC1/MUC4 was clearly better than that of the other patient groups composed of single-positive and double-positive group (p = 0.001). Within 60 months after surgery, the double-positive group showed worse survival than the single-positive group
Fig. 4 Correlation between MUC1/MUC4 co-expression and disease-free survival. The disease-free survival of patients with doublenegative expression of MUC1/MUC4 was statistically better than that of other groups composed of single-positive and double-positive group (p \ 0.0001), whereas MUC1/MUC4 double-positive group showed the poorest disease-free survival
Table 4 Multivariate analysis of overall survival
Table 5 Multivariate analysis of disease-free survival
Variable
HR
95 % CI
p
Tumor location
HR
95 % CI
p
1.10–3.26
0.022*
0.78–2.21
0.28
1.26–4.41
0.0094*
Lymph node metastasis
Tongue, mandible, buccal, and lip
1
Maxilla and floor of mouth
1.68
0.84–3.35
0.13
Pathologic T classification
Negative (N0)
1
Positive (N1–N3)
1.93
Mode of invasion
T1 or T2
1
T3 or T4
1.53
0.71–3.08
0.27
Lymph node metastasis
1–3
1
Diffuse invasive (4c or 4d)
1.33
Vascular invasion
Negative (N0)
1
Positive (N1–N3)
2.24
1.09–4.43
0.029*
Tumor differentiation Well or moderate Poor
Variable
Absent
1
Present
2.43
MUC1/MUC4 expression 1 3.02
1.15–7.06
0.026*
Mode of invasion 1–3
1
Diffuse invasive (4c or 4d)
1.57
Double-negative (-/-) Single-positive (?/- or -/?)
1 2.33
1.20–4.66
0.013*
Double-positive (?/?)
2.71
1.44–5.32
0.0019*
HR hazard ratio, 95 % CI 95 % confidence interval 0.77–3.14
0.21
*
0.92–5.13
0.074
Analysis of risk factors correlated with regional recurrence
Statistically significant
Vascular invasion Absent
1
Present
2.27
MUC1/MUC4 expression Double-negative (-/-)
1
Single-positive (?/- or -/?)
3.49
1.37–9.98
0.008*
Double-positive (?/?)
3.60
1.40–10.42
0.007*
HR hazard ratio, 95 % CI 95 % confidence interval *
Statistically significant
survival. Taken together, double-positivity/negativity of MUC1/MUC4 expression was a powerful predictor for poor/ favorable disease-free survival of patients with OSCC.
Regional recurrence occurred with the increase of MUC1/ MUC4 positivity; this occurred in six of the 86 (7 %) patients in MUC1/MUC4 double-negative group, in 12 of 55 (22 %) patients in single-positive group, and in 18 of 65 (28 %) patients in double-positive group (p = 0.0003, Chisquare test, Table 3). Multivariate analysis of variables correlated with regional recurrence is summarized in Table 6. In multivariate analysis, MUC1/MUC4 coexpression was found to be an independent risk factor (single-positive group, HR 3.29, 95 % CI 1.16–10.19,
123
Int J Clin Oncol Table 6 Multivariate analysis of variables correlated with subsequent lymph node metastasis Variable
HR
95 % CI
p
0.73–3.69
0.22
0.65–5.55
0.23
Mode of invasion 1–3
1
Diffuse invasive (4c or 4d)
1.66
Vascular invasion Absent
1
Present
1.96
MUC1/MUC4 expression Double-negative (-/-)
1
Single-positive (?/- or -/?)
3.29
1.16–10.19
0.025*
Double-positive (?/?)
4.35
1.65–12.90
0.0025*
HR hazard ratio, 95 % CI 95 % confidence interval *
Statistically significant
p = 0.025; double-positive group, HR 4.35, 95 % CI 1.65–12.90, p = 0.0025, respectively). Among all factors evaluated in this study, MUC1/MUC4 double-positivity was considered to be a most strong marker to predict regional recurrence in OSCC patients.
Discussion Currently, the indications for elective neck dissection in clinically N0 patients with OSCC remain controversial. Regional recurrence is one of the main causes of death from OSCC [4, 5]. However, neck dissection often results in serious complications. In this study, our results showed that double-negativity for MUC1/MUC4 strongly correlated with better outcomes after surgery and that MUC1/ MUC4 double-positivity was the strong independent prognostic factor for regional recurrence. Therefore, the co-expression profile of MUC1/MUC4 may be a potent indication for elective neck dissection. Several previous studies have demonstrated the usefulness of MUC1 and MUC4 in cancer detection and prognosis. Implication of mucins in the pathogenesis as oncogenes has provided them as highly attractive targets for cancer therapy [9]. Measurement of circulating MUC1N levels has been used to monitor the clinical course of patients with breast cancer and to detect early disease recurrence [9, 35]. On the other hand, MUC4 was found to mask the antibody-binding epitope of ErbB2 and lead to diminished Herceptin binding, suggesting an interesting attribute of MUC4 that may prove useful in therapeutic planning [36, 37]. The present study demonstrated that MUC1/MUC4 combination was a powerful indicator for tumor aggressiveness and poor outcome in OSCC. The MUC1/MUC4 double-positive group showed the poorest
123
overall and disease-free survival, whereas patients with MUC1/MUC4 double-negative expression showed the most favorable prognosis. Multivariate analysis revealed that MUC1/MUC4 double-positivity was a strong independent prognostic factor for overall and disease-free survival. This is the first report to provide evidence of the clinical prognostic value of the combination of MUC1 and MUC4 expression in OSCC and suggests that determination of the combined MUC1/MUC4 expression status in biopsy specimens has the potential to influence decisions regarding treatment. Furthermore, these mucins may be attractive targets for the therapy in OSCC in the future. Similar to this study, previous immunohistochemical studies of intrahepatic cholangiocarcinoma-mass forming type [10] and gallbladder carcinoma [32] have attempted to reveal combinations of mucins as prognostic indicators. Shibahara et al. [10] have demonstrated the tendency of MUC1/MUC4 co-expression (p = 0.065) and that patients with MUC1/MUC4 double-positive expression show the poorest outcomes (100 % death within 20 months after surgery), whereas patients with MUC1/MUC4 doublenegative expression show better outcomes (p = 0.0004) in intrahepatic cholangiocarcinoma-mass forming type. Furthermore, other immunohistochemical studies using MUC1, MUC2 and MUC4 in gallbladder carcinoma showed a significant correlation between MUC4 and MUC1 expression (p = 0.004), and that patients with a MUC4(?)/MUC2(-) pattern show the poorest outcomes compared with those of patients with other combinations of MUC4/MUC2 expression (p \ 0.05) [32]. In this manner, the combination of several mucin-expression profiles may be used to represent the character of tumors for patient prognosis because mucin expression varies considerably depending on the tissues, organs, and tumors [23]. The biological mechanism of co-expression of MUC1 and MUC4 is unclear at present. To date, epigenetic regulations such as DNA methylation and histone modification have been reported to affect both MUC1 [38] and MUC4 [39, 40] expression, suggesting that the regulatory mechanism of this co-expression may be caused by aberrant status of epigenome. Therefore, comprehensive analysis including mRNA and protein expression and epigenetic aberrations of MUC1 and MUC4 in OSCC would be an interesting area of future study. Moreover, to date, many variants of alternative splicing of the MUC1 transcript have also been observed [41] and some MUC1 splice variants are expressed in human neoplasms such as ovarian tumors [42] and ductal carcinoma in situ of the breast [43]. Interestingly, Obermair et al. [42] have demonstrated that specific MUC1 splice variants are associated with tumor malignancy, suggesting its usefulness as a diagnostic tool. Also, alternative splicing of the MUC4 has been reported [44] and some of these are expressed in pancreatic
Int J Clin Oncol
carcinomas but not in the normal pancreas [45, 46]. Taken together, the dysfunction of these regulations of MUC1 and MUC4 expression such as epigenetic alterations and alternative splicing might be a candidate for a molecular marker and a therapeutic target of OSCC in the future. In conclusion, this study demonstrated significant coexpression of MUC1/MUC4 and that the clinical prognostic value of the combination of MUC1 and MUC4 expression in OSCC. The results suggest that determination of the co-expression status of MUC1/MUC4 in biopsy specimens has the potential to influence decisions regarding treatment, such as the need for elective neck dissection. Acknowledgments The authors thank Mr. Yoshiharu Atsuji and Ms. Yukari Nishimura for their excellent technical assistance. This study was supported in part by Grants-in-Aid for Scientific Research (B) 23390085 to S. Yonezawa; Scientific Research (C) 24590447 to M. Higashi; Scientific Research (B) 23390466 to K. Sugihara and Young Scientists (B) 24792239 to T. Hamada from the Ministry of Education, Science, Sports, Culture and Technology, Japan.
13.
14.
15.
16.
17.
18.
19. Conflict of interest The authors declare that they have no conflicts of interest. 20.
21.
References 1. Jemal A, Bray F, Center MM et al (2011) Global cancer statistics. CA Cancer J Clin 61:69–90 2. Warnakulasuriya S (2009) Global epidemiology of oral and oropharyngeal cancer. Oral Oncol 45:309–316 3. Cheng A, Schmidt BL (2008) Management of the N0 neck in oral squamous cell carcinoma. Oral Maxillofac Surg Clin N Am 20:477–497 4. Kowalski LP, Bagietto R, Lara JR et al (2000) Prognostic significance of the distribution of neck node metastasis from oral carcinoma. Head Neck 22:207–214 5. Capote A, Escorial V, Munoz-Guerra MF et al (2007) Elective neck dissection in early-stage oral squamous cell carcinoma— does it influence recurrence and survival? Head Neck 29:3–11 6. Mahfouz ME, Rodrigo JP, Takes RP et al (2010) Current potential and limitations of molecular diagnostic methods in head and neck cancer. Eur Arch Otorhinolaryngol 267:851–860 7. Hollingsworth MA, Swanson BJ (2004) Mucins in cancer: protection and control of the cell surface. Nat Rev Cancer 4:45–60 8. Moniaux N, Escande F, Porchet N et al (2001) Structural organization and classification of the human mucin genes. Front Biosci 6:1192–1206 9. Kufe DW (2009) Mucins in cancer: function, prognosis and therapy. Nat Rev Cancer 9:874–885 10. Shibahara H, Tamada S, Higashi M et al (2004) MUC4 is a novel prognostic factor of intrahepatic cholangiocarcinoma-mass forming type. Hepatology 39:220–229 11. Higashi M, Yonezawa S, Ho JJ et al (1999) Expression of MUC1 and MUC2 mucin antigens in intrahepatic bile duct tumors: its relationship with a new morphological classification of cholangiocarcinoma. Hepatology 30:1347–1355 12. Nakamura A, Horinouchi M, Goto M et al (2002) New classification of pancreatic intraductal papillary-mucinous tumour by
22.
23.
24.
25. 26.
27.
28.
29.
30.
31.
mucin expression: its relationship with potential for malignancy. J Pathol 197:201–210 Kitazono I, Higashi M, Kitamoto S et al (2013) Expression of MUC4 mucin is observed mainly in the intestinal type of intraductal papillary mucinous neoplasm of the pancreas. Pancreas 42:1120–1128 Osako M, Yonezawa S, Siddiki B et al (1993) Immunohistochemical study of mucin carbohydrates and core proteins in human pancreatic tumors. Cancer 71:2191–2199 Sagara M, Yonezawa S, Nagata K et al (1999) Expression of mucin 1 (MUC1) in esophageal squamous-cell carcinoma: its relationship with prognosis. Int J Cancer 84:251–257 Takao S, Uchikura K, Yonezawa S et al (1999) Mucin core protein expression in extrahepatic bile duct carcinoma is associated with metastases to the liver and poor prognosis. Cancer 86:1966–1975 Saitou M, Goto M, Horinouchi M et al (2005) MUC4 expression is a novel prognostic factor in patients with invasive ductal carcinoma of the pancreas. J Clin Pathol 58:845–852 Shibahara H, Tamada S, Goto M et al (2004) Pathologic features of mucin-producing bile duct tumors: two histopathologic categories as counterparts of pancreatic intraductal papillary-mucinous neoplasms. Am J Surg Pathol 28:327–338 Tamada S, Goto M, Nomoto M et al (2002) Expression of MUC1 and MUC2 mucins in extrahepatic bile duct carcinomas: its relationship with tumor progression and prognosis. Pathol Int 52:713–723 Tamada S, Shibahara H, Higashi M et al (2006) MUC4 is a novel prognostic factor of extrahepatic bile duct carcinoma. Clin Cancer Res 12:4257–4264 Tamura Y, Higashi M, Kitamoto S et al (2012) MUC4 and MUC1 expression in adenocarcinoma of the stomach correlates with vessel invasion and lymph node metastasis: an immunohistochemical study of early gastric cancer. PLoS One 7:e49251 Tsutsumida H, Goto M, Kitajima S et al (2007) MUC4 expression correlates with poor prognosis in small-sized lung adenocarcinoma. Lung Cancer 55:195–203 Yonezawa S, Goto M, Yamada N et al (2008) Expression profiles of MUC1, MUC2, and MUC4 mucins in human neoplasms and their relationship with biological behavior. Proteomics 8:3329–3341 Yonezawa S, Higashi M, Yamada N et al (2010) Significance of mucin expression in pancreatobiliary neoplasms. J Hepatobiliary Pancreat Sci 17:108–124 Yonezawa S, Higashi M, Yamada N et al (2008) Precursor lesions of pancreatic cancer. Gut Liver 2:137–154 Yonezawa S, Taira M, Osako M et al (1998) MUC-1 mucin expression in invasive areas of intraductal papillary mucinous tumors of the pancreas. Pathol Int 48:319–322 Yonezawa S, Higashi M, Yamada N et al (2011) Mucins in human neoplasms: clinical pathology, gene expression and diagnostic application. Pathol Int 61:697–716 Yonezawa S, Sato E (1997) Expression of mucin antigens in human cancers and its relationship with malignancy potential. Pathol Int 47:813–830 Yonezawa S, Nakamura A, Horinouchi M et al (2002) The expression of several types of mucin is related to the biological behavior of pancreatic neoplasms. J Hepatobiliary Pancreat Surg 9:328–341 Hamada T, Nomura M, Kamikawa Y et al (2012) DF3 epitope expression on MUC1 mucin is associated with tumor aggressiveness, subsequent lymph node metastasis, and poor prognosis in patients with oral squamous cell carcinoma. Cancer 118:5251–5264 Hamada T, Wakamatsu T, Miyahara M et al (2012) MUC4: a novel prognostic factor of oral squamous cell carcinoma. Int J Cancer 130:1768–1776
123
Int J Clin Oncol 32. Lee HK, Cho MS, Kim TH (2012) Prognostic significance of muc4 expression in gallbladder carcinoma. World J Surg Oncol 10:224 33. Sobin LGM, Wittekind Ce (2009) TNM Classification of malignant tumors. In: Hoboken NJ (ed), 7th ed John Wiley & Sons, Inc. 34. Yamamoto E, Kohama G, Sunakawa H et al (1983) Mode of invasion, bleomycin sensitivity, and clinical course in squamous cell carcinoma of the oral cavity. Cancer 51:2175–2180 35. Hayes DF, Sekine H, Ohno T et al (1985) Use of a murine monoclonal antibody for detection of circulating plasma DF3 antigen levels in breast cancer patients. J Clin Invest 75:1671–1678 36. Nagy P, Friedla¨nder E, Tanner M et al (2005) Decreased accessibility and lack of activation of ErbB2 in JIMT-1, a Herceptinresistant, MUC4-expressing breast cancer cell line. Cancer Res 65:473–482 37. Singh AP, Chaturvedi P, Batra SK (2007) Emerging roles of MUC4 in cancer: a novel target for diagnosis and therapy. Cancer Res 67:433–436 38. Yamada N, Nishida Y, Tsutsumida H et al (2008) MUC1 expression is regulated by DNA methylation and histone H3 lysine 9 modification in cancer cells. Cancer Res 68:2708–2716 39. Yamada N, Nishida Y, Tsutsumida H et al (2009) Promoter CpG methylation in cancer cells contributes to the regulation of MUC4. Br J Cancer 100:344–351
123
40. Vincent A, Ducourouble MP, Van Seuningen I (2008) Epigenetic regulation of the human mucin gene MUC4 in epithelial cancer cell lines involves both DNA methylation and histone modifications mediated by DNA methyltransferases and histone deacetylases. FASEB J 22:3035–3045 41. Zhang L, Vlad A, Milcarek C et al (2013) Human mucin MUC1 RNA undergoes different types of alternative splicing resulting in multiple isoforms. Cancer Immunol Immunother 62:423–435 42. Obermair A, Schmid BC, Packer LM et al (2002) Expression of MUC1 splice variants in benign and malignant ovarian tumours. Int J Cancer 100:166–171 43. Schmid BC, Rudas M, Fabjani G et al (2003) Evaluation of MUC1 splice variants as prognostic markers in patients with ductal carcinoma in situ of the breast. Oncol Rep 10:1981–1985 44. Choudhury A, Moniaux N, Winpenny JP et al (2000) Human MUC4 mucin cDNA and its variants in pancreatic carcinoma. J Biochem 128:233–243 45. Moniaux N, Escande F, Batra SK et al (2000) Alternative splicing generates a family of putative secreted and membrane-associated MUC4 mucins. Eur J Biochem 267:4536–4544 46. Choudhury A, Moniaux N, Ringel J et al (2001) Alternate splicing at the 30-end of the human pancreatic tumor-associated mucin MUC4 cDNA. Teratog Carcinog Mutagen 21:83–96
