• Aucun résultat trouvé

Motility-related protein 1 (MRP-1/CD9) expression in urothelial bladder carcinoma and its relation to tumor recurrence and progression

N/A
N/A
Protected

Academic year: 2022

Partager "Motility-related protein 1 (MRP-1/CD9) expression in urothelial bladder carcinoma and its relation to tumor recurrence and progression"

Copied!
10
0
0

Texte intégral

(1)

Article

Reference

Motility-related protein 1 (MRP-1/CD9) expression in urothelial bladder carcinoma and its relation to tumor recurrence and

progression

MHAWECH-FAUCEGLIA, Paulette, et al .

Abstract

BACKGROUND: CD9 has been implicated in cell adhesion, motility, and proliferation, and numerous studies have demonstrated its prognostic value in different solid tumors. The objective of this study was to determine the relation of CD9 expression to tumor grade and tumor stage of urothelial carcinoma of the bladder and to define the value of CD9 in predicting the behavior of superficial papillary tumors (SPTs) (pathologic Ta [pTa] and pT1). METHODS:

Three hundred twenty patients (118 patients with pTa tumors, 111 patients with pT1 tumors, and 91 patients with pT2 tumors) were examined for CD9 expression using immunohistochemistry applied on formalin fixed, paraffin embedded tissue. Patients were stratified into 3 categories, depending on CD9 expression: positive (> 50% positive cells), reduced (5-50% positive cells), or negative (< 5% positive cells). RESULTS: Loss of CD9 expression was found to be associated significantly with high-grade and high-stage urothelial tumors (P < 0.0001). A reduced/negative (altered) CD9 expression was associated with SPT progression, but not with recurrence (P < 0.001). Patients who had pTa [...]

MHAWECH-FAUCEGLIA, Paulette, et al . Motility-related protein 1 (MRP-1/CD9) expression in urothelial bladder carcinoma and its relation to tumor recurrence and progression. Cancer , 2003, vol. 98, no. 8, p. 1649-1657

PMID : 14534881

DOI : 10.1002/cncr.11698

Available at:

http://archive-ouverte.unige.ch/unige:28149

Disclaimer: layout of this document may differ from the published version.

1 / 1

(2)

Motility-Related Protein 1 (MRP-1/CD9) Expression in Urothelial Bladder Carcinoma and its Relation to

Tumor Recurrence and Progression

Paulette Mhawech,M.D.1

Franc¸oı´s Herrmann,M.D.2

Monique Coassin,M.T.1

Louis Guillou,M.D.3

Christophe E. Iselin,M.D.4

1Department of Pathology, Geneva University Hos- pital, Geneva, Switzerland.

2Department of Geriatrics, Geneva University Hos- pital, Geneva, Switzerland.

3Department of Pathology, Lausanne University Hospital, Lausanne, Switzerland.

4Urology Clinic, Geneva University Hospital, Ge- neva, Switzerland.

The authors thank Mr. Ludovic Metral for his as- sistance in searching the archives, Ms. Danielle Fontana for her technical assistance, and the staff of the immunohistochemical laboratory for their help. The authors also thank Dr. Sophia Taylor for her critical review of the article.

Address for reprints: Paulette Mhawech, M.D., Pa- thologie Clinique, Geneva University Hospital, CMU, Michel-Servet, 1, 1211 Geneva, 4 Switzerland;

Fax: (011) 41-223724920; E-mail: paulette.

mhawech-fauceglia@hcuge.ch

Received April 1, 2003; revision received July 11, 2003; accepted July 15, 2003.

BACKGROUND.CD9 has been implicated in cell adhesion, motility, and prolifera- tion, and numerous studies have demonstrated its prognostic value in different solid tumors. The objective of this study was to determine the relation of CD9 expression to tumor grade and tumor stage of urothelial carcinoma of the bladder and to define the value of CD9 in predicting the behavior of superficial papillary tumors (SPTs) (pathologic Ta [pTa] and pT1).

METHODS. Three hundred twenty patients (118 patients with pTa tumors, 111 patients with pT1 tumors, and 91 patients with pT2 tumors) were examined for CD9 expression using immunohistochemistry applied on formalin fixed, paraffin embedded tissue. Patients were stratified into 3 categories, depending on CD9 expression: positive (⬎ 50% positive cells), reduced (5–50% positive cells), or negative (⬍5% positive cells).

RESULTS.Loss of CD9 expression was found to be associated significantly with high-grade and high-stage urothelial tumors (P 0.0001). A reduced/negative (altered) CD9 expression was associated with SPT progression, but not with recur- rence (P0.001). Patients who had pTa or pT1 tumors with altered CD9 expres- sion had a relative risk of 5.59 (P 0.005; 95% confidence interval [95% CI], 1.69 –18.48) for progression compared with patients who had tumors with positive CD9 expression. Kaplan–Meier curves showed that a lack of CD9 expression was associated significantly with progression free survival (P0.001; log-rank test), but not with recurrence. In patients with SPTs, multivariate Cox proportional hazards regression analysis revealed that negative CD9 expression was an independent prognostic marker for the prediction of tumor progression (P0.007; 95% CI, 0.11– 0.70).

CONCLUSIONS.In patients with urothelial bladder carcinoma, CD9 expression was associated significantly with tumor stage and grade, and a loss of CD9 expression was an independent prognostic factor for predicting progression in patients with SPTs. Thus, CD9 immunoexpression is a potential new predictor of tumor behavior in patients with SPTs of the urinary bladder.Cancer2003;98:1649 –57.

© 2003 American Cancer Society.

KEYWORDS: CD9 expression, superficial papillary tumors, tumor behavior, urinary bladder.

B

ladder carcinoma is the fourth leading cause of cancer in men and the eighth leading cause in women, with almost 54,000 new diagnoses registered each year in the U.S.1 Most urothelial bladder carcinomas (75%) present as superficial tumors.2Two distinct forms ofsuperficial papillary tumors (SPTs) may exist. The first form (Ta) includes noninvasive tumors with a high risk of recurrence (50 –75%) and a low risk of progression (5%). The second form (T1) includes

© 2003 American Cancer Society DOI 10.1002/cncr.11698

(3)

tumors that invade the lamina propria, but not the muscularis propria, and carry a high likelihood of pro- gression (30 –50%).2,3Although most recurrent super- ficial tumors are not life-threatening, they require fre- quent surveillance, and a significant number of patients also may need additional treatment, such as endoscopic resection and intravesical chemotherapy installation. Conversely, a small number of superficial tumors are life-threatening; these need to be detected early, so that adequate aggressive therapy may be instituted.4,5In this context, identifying and develop- ing a marker capable of predicting tumor behavior is of great importance. Although a large number of markers, such as oncogenes, tumor suppressor genes, growth factor receptors, adhesion molecules, and cell proliferation markers, have been evaluated extensively as potential prognostic factors in patients with bladder carcinoma,6 –9tumor grade and stage still are consid- ered the most significant variables for predicting pa- tient outcome.

The transmembrane-4 superfamily (TM4SF) is a surface glycoprotein family that is involved in cell growth, adhesion, and motility. It is comprised of 20 known members with 20 –30% sequence similarity, five of among which (motility-related protein 1 [MRP1/CD9], ME491/CD63, KAI1/CD82, CD151, and CD81] may be implicated in cell migration, prolifera- tion, and tumor cell metastasis.10,11TM4SF has a basic structure of four hydrophobic transmembrane do- mains (TM1–TM4), two extracellular loops bearing the site of gylcosylation, and short cytoplasmic N and C termini.12The motility-related protein/CD9 is located on chromosome 12 (12p13) and is distributed widely among all cell types. To date, CD9 is the best charac- terized of the TM4SF members. It is capable of inter- acting with other transmembrane proteins, such as integrins and other tetraspanins, to form a complex that facilitates cell adhesion, motility, and signal- ing.13,14 CD9 has an important role in inhibiting cell motility in numerous neoplastic cell lines, including lung, gastric, pancreatic, and bladder carcinomas.11,15 Clinical studies have shown the important prognostic value of CD9 in adenocarcinoma of the lung, colon, breast, pancreatic, and in squamous cell carcinoma of the esophagus and the oral cavity.16 –22However, be- cause the value of CD9 protein as a prognostic factor for urothelial bladder carcinoma has not been exam- ined to date, we attempted to define its value in the current study. First, we examined the relation between CD9 expression and the two most significant estab- lished prognostic factors, grade and stage, in urothe- lial bladder carcinoma. In a second step, we examined the potential value of CD9 in predicting tumor recur-

rence and tumor progression in patients with pTa and pT1 disease.

MATERIALS AND METHODS Patient Population

A retrospective, consecutive study covering an 11-year period (1988 –1999) was conducted. Patients with a prior diagnosis of bladder carcinoma with or without treatment were excluded from the study. Pathologic material included transurethral resection specimens from bladder tumors. For every patient, hematoxylin and eosin stained slides were reviewed by one pathol- ogist (P.M.), and the histologic grade and stage were reassessed. The 1973 World Health Organization grad- ing system was used for histologic assessment, and the sixth edition of the TNM classification system was used for staging disease.23,24Follow-up data were re- trieved from our archives and from the medical records. The closing date for the study was the date of the patient’s last histologic diagnosis, cystectomy, or death with autopsy performed. Treatment consisted of endoscopic resection for superficial tumors (pTa and pT1) and intravesical bacillus Calmette–Guerin (BCG) for patients with high-risk tumors, such as high-grade lesions, and frequent recurrences. Depending on fol- low-up, tumors were stratified into four subgroups: 1) a tumor regression subgroup (from pTa to no detect- able tumor and from pT1 to no detectable tumor or to pTa); 2) a tumor recurrence subgroup (from pTa to pTa and from pT1 to pT1); 3) a tumor progression subgroup (from pTa toⱖpT1 and from pT1 toⱖpT2);

and 4) a subgroup of tumors with no follow-up. To address the first issue of the study, i.e., the relation between CD9 expression and tumor stage and grade, pT2 tumors were included in the analysis as well as pTa and pT1 tumors. To address the second issue of the study, i.e., the relation of CD9 immunoexpression with both tumor recurrence and tumor progression in patients with SPTs of the bladder, only pTa and pT1 tumors were included in the analysis.

Immunohistochemistry

Immunohistochemistry (IHC) was performed on par- affin embedded tissues. Tissue sections were deparaf- finized with xylene and washed with ethanol. For im- munostaining enhancement, sections were pretreated in a microwave oven in 0.01 M citrate buffer, pH 6.0, at 98 °C for 30 minutes. Sections were incubated with a monoclonal anti-CD9 antibody (Novacastra, New- castle, U.K.) diluted at 1:20. Endogenous peroxidase was blocked with 0.3% hydrogen peroxidase for 5 min- utes. Thus, sections were incubated with mouse Envi- sion horseradish peroxidase for 30 minutes. IHC was performed using the automated stainer, Dako Au- 1650 CANCER October 15, 2003 / Volume 98 / Number 8

(4)

tostainer (Dakopatts, Copenhagen, Denmark). These incubations were performed at room temperature, and sections were washed in Tris buffered saline be- tween incubations. Diaminobenzidine complex was used as the chromogen. Tissue sections from patients with fibroadenoma were used as positive controls.

Staining of normal bladder mucosa with anti-CD9 showed strong membranous positivity of the urothe- lium (Fig. 1). In negative controls, a normal goat se- rum was used in place of the primary antibody, result- ing in a lack of detectable staining.

IHC for CD9 staining was performed on the first resection material only; therefore, subsequent biop- sies were not stained. Evaluation of the IHC was done twice by 1 pathologist (P. M.) with a 1-month interval.

All of the tissue on the slide was scanned for CD9 expression, and the average expression was calculated for the entire slide. Samples for CD9 expression were stratified into three categories, depending on the percentage of CD9 positive cells, as described previ-

ously.21Samples with⬎50% CD9 positive tumor cells were categorized withpositive CD9 expression, sam- ples with 5–50% stained tumor cells were categorized withreducedCD9 expression, and samples with⬎5%

stained tumor cells were categorized with negative CD9 expression.

Statistical Analysis

For statistical analysis, the following baseline variables were considered for their prognostic value: age at pre- sentation, gender, tumor stage, tumor grade, and CD9 expression (considered as a categoric variable). Asso- ciations between variables were studied using the chi- square test or the Fisher exact test, as appropriate. The TABLE 1

Clinical and Histologic Data of the Patients Analyzed

Characteristic No. of patients

Pathologic T classification

pTa 118

pT1 111

pT2 91

Tumor grade

Grade 1 85

Grade 2 122

Grade 3 113

Gender (male:female ratio) 257:63 (4.1) Age (yrs)

MeanSD 70.411.7

Range 27–98

Follow-up (mos)

Mean 30

Median 43.6

Range 12–240

First progression in stage

pTa to pT1 or pT2 28

pT1 to pT2 32

Regression in stage

pTa to no tumor 30

pT1 to pTa or no tumor 33

Recurrences

pTa 58

pT1 29

Death from disease

pTa 1

pT1 0

pT2 2

Death from unrelated causes

pTa 6

pT1 5

pT2 2

No follow-up

pTa 9

pT1 19

pT2 87

Metastasis 10

SD: standard deviation.

FIGURE 1.Using immunohistochemistry in a section from normal bladder wall, strong positivity of the normal urothelium is seen in a membranous pattern when stained with anti-CD9 protein.

(5)

date of diagnosis was considered the time of origin.

For recurrence free and progression free intervals, we considered the first recurrence and the first progres- sion, respectively, as events. Multivariate logistic re- gression models were defined to predict the occur- rence of the first progression or the first recurrence while adjusting for all the baseline variables. Recur- rence free and progression free survival curves were computed using the Kaplan–Meier method. For com- parison between curves, we used the log-rank test.

Multivariate analyses based on a Cox proportional hazards model were used to identify the most signifi- cant factors related to outcome. The results of the multivariate analyses are expressed in terms of either odds ratios or hazard ratios, which were derived from the estimated regression coefficients along with their 95% confidence intervals (95% CI). P values ⱕ 0.05 were considered statistically significant. All analysis were performed with the Stata statistical software package (version 7.0; Stata Corporation, College Sta- tion, TX).

RESULTS

Three hundred twenty patients met our inclusion cri- teria. Clinical data, grade, and pathologic stage are summarized in Table 1. Of 320 tumors, 118 tumors were classified as pTa, 111 tumors were classified as pT1, and 91 tumors were classified as pT2 at initial diagnosis. Eighty-five tumors were well differentiated (Grade 1), 122 tumors were moderately differentiated (Grade 2), and 113 tumors were poorly differentiated (Grade 3). The age of patients ranged from 27 years to 98 years (mean, 70.4 years). There were 257 males and 63 females, with a male-to-female ratio of 4:1. The follow-up period ranged from 12 months to 240 months (median, 30 months; mean, 43.6 months). For 202 patients, the follow-up period wasⱖ12 months.

Shorter follow-up periods of 2 months, 3 months, and 6 months were noted in 3 patients (1 patient each with pTa, 1 pT1, and 1 pT2 tumors, respectively) because they underwent a cystectomy shortly after the biopsy.

Of the 115 patients with no follow-up available, there were 87 patients with pT2 tumors, 9 patients with pTa tumors, and 19 patients with pT1 tumors. Of 118 pTa tumors, 30 tumors regressed, 58 tumors recurred, and 28 tumors progressed. Of 111 pT1 tumors, 33 tumors regressed (8 to pTa and 25 to no detectable tumor), 29 tumors recurred, and 32 tumors progressed toⱖpT2.

Three patients died from disease-related causes, and 13 patients died from unrelated causes. Ten patients developed metastases during the course of their dis- ease. With regard to CD9 expression, 164 tumors were positive, 82 tumors had reduced expression, and 74

tumors were negative (10 tumors that had weak mem- branous staining were included in the negative group).

A correlation was observed between tumor stage and grade, as expected, with aPvalue⬍0.0001 (Table 2). Patients with well differentiated tumors (Grade 1) presented more often at an earlier stage (pTa; 66%), and patients with poorly differentiated tumors (Grade 3) presented more often at a more advanced stage (pT2; 84%). Patients with moderately differentiated tumors (Grade 2) presented more often as pT1 tumors (63%).

The relation between CD9 expression and tumor stage and grade is noted in Table 3 (Figs. 2A,B, 3A,B).

A significant association between CD9 expression and tumor stage was noted (P ⬍ 0.0001). Among CD9 positive tumors, 60% were pTa, 32% were pT1, and 7%

were pT2. Among CD9 negative tumors, 3% were pTa, 27% were pT1, and 70% were pT2. Reduced expression of CD9 was seen in 20%, 46%, and 32% of pTa, pT1, and pT2 tumors, respectively. There also was a signif- icant association between CD9 expression and tumor grade (P ⬍ 0.0001) (Table 3). Among CD9 positive tumors, 50%, 43%, and 6.7% were Grade 1, Grade 2, and Grade 3, respectively. Among CD9 negative tu- mors, zero tumors were Grade 1, 13.5% of tumors were TABLE 2

Association between Tumor Grade and Pathologic T Classification

Grade

pTa pT1 pT2

Total Pvalue

No. % No. % No. %

Grade 1 78 66.1 7 6.3 0 0.0 85 0.0001

Grade 2 38 32.2 70 63.1 14 15.4 22

Grade 3 2 1.7 34 30.6 77 84.6 113

Total 118 111 91 320

TABLE 3

CD9 Expression in Relation to Pathologic T Classification and Tumor Grade

Characteristic

CD9 expression

Pvalue Positive Reduced Negative

Total

No. % No. % No. %

Pathologic T classification

pTa 99 60.4 17 20.7 2 2.7 117 0.0001

pT1 53 32.3 38 46.3 20 27.0 109

pT2 12 7.3 27 32.9 52 70.3 91

Tumor grade

1 82 50.0 3 3.7 0 0.0 84 0.0001

2 71 43.3 41 50.0 10 13.5 120

3 11 6.7 38 46.3 64 86.5 113

Total 164 82 74

1652 CANCER October 15, 2003 / Volume 98 / Number 8

(6)

Grade 2, and 86.5% of tumors were and Grade 3.

Reduced expression of CD9 was observed more fre- quently in Grade 2 tumors (50%) and Grade 3 tumors (46.3%).

The relation between CD9 expression and tumor behavior is illustrated in Table 4. Applying logistic regression, the crude odds ratio associated with posi- tive CD9 expression (which is an estimate of the rela- tive risk of progression vs. tumor regression) amounted to 0.15. This means that patients who had pTa or pT1 tumors with⬍5% CD9 positive cells were 6.72 (1/0.15) times more likely to develop tumor pro-

gression (P ⱕ 0.001; 95% CI, 2.55–17.71) compared with patients who had tumors with ⬎ 50% CD9 ex- pression. Adjusted for age, gender, and grade, the odds ratio is 5.59 (P ⫽ 0.005; 95% CI, 1.69 –18.48).

Thus, when adjusting for all other variables, aberrant CD9 expression remained a significant predictor of tumor progression (Table 5).

The Kaplan–Meier curves using the log-rank test were performed to evaluate the time delay to the first recurrence and to the first progression in patients with pTa and pT1 tumors (Figs. 4, 5). The Kaplan–Meier curves showed that CD9 expression was not signifi- FIGURE 3.(A) A poorly differentiated,

Grade 3, urothelial bladder carcinoma with invasion of the muscularis propria (pathologic T2) stained with hematoxylin and eosin. (B) Using immunohistochem- istry, a total loss of CD9 reactivity is seen in tumor cells (5% of tumor cells are positive).

FIGURE 2. (A) A well differentiated, Grade 1, noninvasive (pathologic Ta), papillary urothelial bladder carcinoma stained with hematoxylin and eosin staining. (B) Using immunohistochemis- try, the tumor cells show strong mem- branous reactivity for the CD9 antibody (90% of tumor cells are positive).

(7)

cant in predicting recurrence free survival in patients with pTa or pT1 tumors (P⫽0.33). However, positive CD9 expression was associated significantly with slower disease progression in patients with pTa and pT1 tumors (P⬍ 0.001). Thus, patients who had tu- mors with positive CD9 expression were more likely to have a longer time to progression compared with pa- tients who had tumors with negative and reduced CD9 expression.

A multivariate Cox regression analysis is shown in Table 6. The Cox regression model revealed that only the pathologic grade was statistically significant (P

⫽0.003) in predicting recurrence. Thus, only patho- logic grade appeared to be an independent factor for predicting tumor recurrence, and CD9 did not add any information. The same Cox regression model applied to study the predictors of tumor progression revealed that pathologic grade, stage, and CD9 expression (pos- itive vs. negative) were statistically significant (P

⬍0.0001). Most important, when simultaneously ad- justing for all of the other variables, only CD9 expres- sion maintained its independent prognostic value for predicting tumor progression (P ⫽ 0.007), whereas pathologic grade and stage failed to maintain any value.

DISCUSSION

The current study used a large data base of 320 pa- tients to address two issues. To address the first issue (i.e., the relation of CD9 expression to tumor grade and stage), all 320 patients were included. However, to address the second issue (i.e., the relation of CD9 to

tumor behavior in SPTs), data from patients with pT2 tumors were not relevant and were not included.

Thus, the study included 202 patients with SPTs who had long-term follow-up (range, 12–240 months). This long follow-up provided an accurate and complete understanding of the behavior of these repeatedly re- current superficial tumors.

This study showed a strong association of aber- rant CD9 expression (reduced or negative) with two major, well established predictive factors of tumor behavior: tumor grade and stage. Low-grade tumors, which often are positive for CD9 expression, are less likely to become more aggressive and to invade the muscularis propria. Consequently, high-grade tumors, which usually are negative for CD9 expression, are more likely to progress and invade.

With regard to tumor behavior, only the extreme values of CD9 were important in predicting tumor progression in patients with SPTs. That is, patients who had pTa and pT1 tumors with an absence of CD9 expression were more likely to progress compared with patients who had tumors with positive CD9 ex- pression. Thus, CD9 expression may be used as an additional marker for predicting disease progression in both pTa and pT1 superficial tumors. In patients with these tumors, progression is a life-threatening issue and is worth identifying as such, so that patients can be managed more closely and may be treated more aggressively. In this context, CD9 may be con- sidered a useful tool for identifying patients who are at risk for disease progression and for whom closer fol- low-up is required. When other variables, such as age, gender, grade, stage, and CD9 expression, were taken into consideration, CD9 expression appeared to be a powerful, independent prognostic factor regarding progression free survival. Thus, it appears that CD9 expression was superior to tumor grade and stage in predicting progression free survival. For the probabil- ity of recurrence free survival in patients with SPTs, only tumor grade proved to be an independent pre- dictor, whereas CD9 expression did not seem to have any prognostic value. Although BCG therapy may have an impact on tumor recurrence, there is no evidence that it has an effect on long-term progression in pa- tients with SPTs.25Thus, it is unlikely that BCG was a source of bias in the current study.

This study also showed a strong correlation be- tween tumor grade and depth of invasion. The major- ity of noninvasive tumors were well differentiated, and the invasive tumors were poorly differentiated. Con- versely, pT1 tumors were mostly moderately differen- tiated.5 These findings are in accordance with data from the literature.26 –29

Clinical studies have demonstrated the prognostic TABLE 4

Correlation of CD9 Expression with Tumor Behavior

Tumor behavior

CD9 expression

Pvaluea Positive

Reduced or negative

Total

No. % No. %

Regression

pTa 29 30.20 1 5.0 30 0.041

pT1 23 47.91 10 21.7 33 0.003

First progression

pTa 16 16.66 12 60.0 28 0.001

pT1 8 16.66 24 52.2 32 0.004

Recurrence

pTa 51 53.12 7 35.0 58 0.318

pT1 17 35.41 12 26.1 29 0.199

Total

pTa 96 20 116

pT1 48 46 94

aFisher exact test.

1654 CANCER October 15, 2003 / Volume 98 / Number 8

(8)

value of CD9 in patients with adenocarcinoma of the colon, lung, pancreas, and breast and patients with squamous cell carcinoma of the esophagus and oral cavity.16 –22Miyake et al.16evaluated the value of CD9 expression in breast carcinoma for predicting progno- sis after surgery. In their study, those authors con- cluded that, in addition to hormone receptor status and lymph node status, positive expression of CD9 by tumor cells was correlated with disease free survival and, thus, that testing for CD9 expression could iden- tify patients who may have a high risk for recurrence.

Along the same lines, Huang et al.17showed that pa- tients with breast carcinoma who had negative CD9 expression had shorter disease free survival and a shorter 5-year survival rate compared with patients who had positive CD9 expression. Higashiyama et al.20 found that positive expression of CD9 was an inde- pendent prognostic factor for disease free survival in

patients with adenocarcinoma of the lung. Uchida et al.21found that reduction or absence of CD9 expres- sion correlated well with lymph node metastasis and may predict prognosis in patients with esophageal squamous cell carcinoma. Kusukawa et al.22 found that loss of CD9 staining was associated with a high incidence of lymph node metastasis and a poorer prognosis in patients with squamous cell carcinoma of the oral cavity. However, the value of CD9 in patients with urothelial bladder carcinoma has not been doc- umented to date, and the current study is the first to do so. The lack of studies in the literature on the exact topic discussed in our study make the comparison of our data with others difficult. In addition, because superficial bladder tumors and solid tumors, such as tumors of the colon, breast, and lung, have different behavior (for example, the former are prone to recur TABLE 5

Logistic Regression Analysis of the Prediction of First Recurrence and First Progression among Patients with Pathologic Ta and T1 Tumors

Variable

Recurrence Progression

Crude

OR Pvalue 95% CI

Adjusted

OR Pvalue 95% CI

Crude

OR Pvalue 95% CI

Adjusted

OR Pvalue 95% CI

CD9 expression

Reduced vs. negative 1.22 0.744 0.37–4.07 1.22 0.760 0.34–4.47 0.60 0.325 0.22–1.66 0.63 0.390 0.22–1.82 Positive vs. negative 2.71 0.086 0.87–8.45 3.28 0.113 0.75–14.40 0.15 0.000 0.06–0.39 0.18 0.005 0.05–0.59 Pathologic grade 0.74 0.278 0.43–1.27 2.19 0.095 0.87–5.52 2.31 0.000 1.48–3.62 1.74 0.112 0.88–3.47 Pathologic stage 0.31 0.006 0.14–0.72 0.26 0.014 0.09–0.76 1.40 0.271 0.77–2.53 0.46 0.077 0.20–1.09 Age 0.97 0.076 0.93–1.00 0.97 0.153 0.94–1.01 1.00 0.976 0.97–1.03 1.00 0.980 0.97–1.03 Gender 1.84 0.180 0.75–4.51 1.48 0.427 0.56–3.94 1.50 0.346 0.65–3.49 1.20 0.693 0.48–3.04 OR: odds ratio; 95% CI: 95% confidence interval.

FIGURE 4.Kaplan–Meier estimates of the probability of disease recurrence show that CD9 expression was not significant (negative expression, 15 events in 21 patients; reduced expression, 42 events in 54 patients; positive expres- sion, 128 events in 148 patients) in predicting a delay between the first biopsy and the first disease recurrence in patients with pathologic Ta and T1 tumors (log-rank test;P0.33).

FIGURE 5.Kaplan–Meier estimates of the probability of disease progression show that CD9 expression was significant (negative expression, 12 events in 21 patients; reduced expression, 24 events in 54 patients; positive expression, 24 events in 148 patients) in predicting a delay between the first biopsy and the first sign of disease progression in patients with pathologic Ta and T1 tumors (log-rank test; P 0.001). Vertical tick marks on the curves represent censoring.

(9)

and progress, and the latter disseminate to lymph node and distant metastasis, leading to death), a com- parison of our data with data from patients with these forms of cancer would not be valid. The next step will be to evaluate CD9 expression in muscular invasive tumors (pT2) and its relation to lymph node metasta- sis and survival.

In the current study, aberrant expression of CD9 was found to be associated significantly with tumor grade and depth of invasion in urothelial bladder car- cinoma. In addition, CD9 expression appears to be an independent factor for predicting tumor progression.

Thus, IHC assessment of CD9 expression in superficial tumors (pTa and pT1) may become a promising tool for predicting tumor behavior in patients with super- ficial bladder carcinoma. By selecting those with tu- mors that are likely to progress, patients may benefit from more efficient management earlier in the course of their disease.

REFERENCES

1. Greenlee RT, Hill-Harmon MB, Murray T, Thun M. Cancer statistics, 2001.CA Cancer J Clin.2001;51:15–36.

2. Lee R, Droller MJ. The natural history of bladder cancer;

implications for therapy.Urol Clin North Am.2000;27:1–13.

3. Heney NM. Natural history of bladder cancer: prognostic features and long-term disease course.Urol Clin North Am.

1992;19:429 – 433.

4. Amling CL. Diagnosis and management of superficial blad- der cancer.Curr Probl Cancer.2001;25:219 –278.

5. Soloway MS, Sofer M, Vaidya A. Contemporary management of Stage T1 transitional cell carcinoma of the bladder.J Urol.

2002;167:1573–1583.

6. Knowles MA. What we could do now: molecular pathology of bladder cancer.J Clin Pathol.2001;54:215–221.

7. Desai S, Lim SD, Jimenez RE, et al. Relationship of cytoker- atin 20 and CD44 protein expression with WHO/ISUP grade in pTa and pT1 papillary urothelial neoplasia.Mod Pathol.

2000;13:1315–1323.

8. Sun W, Herrera GA. E-cadherin expression in urothelial carcinoma in situ, superficial papillary transitional cell car- cinoma, and invasive transitional cell carcinoma. Hum Pathol.2002;33:996 –1000.

9. Holmang S, Hedelin H, Anderstrom C, Holmberg E, Busch C, Johansson SL. Recurrence and progression in low grade papillary urothelial tumors.J Urol.1999;162:702–707.

10. Ikeyama S, Koyama M, Yamaoko M, Sasada R, Miyake M.

Suppression of cell motility and metastasis with human motility-related protein (MRP-1/CD9) DNA. J Exp Med.

1993;177:1231–1237.

11. Miyake M, Koyoma M, Seno M, Ikeyama S. Identification of the motility-related protein (MRP-1), recognized by mono- clonal antibody M31-15, which inhibits cell motility.J Exp Med.1991;174:1347–1354.

12. Wright MD, Tomlinson MG. The ins and outs of the trans- membrane 4 superfamily.Immunol Today.1994;15:588 –594.

13. Hemler ME, Mannion BA, Berditchevski F. Association of TM4SF proteins with integrins: relevance to cancer.Biochim Biophys Acta.1996;1297:67–71.

14. Maecker HT, Todd SC, Levy S. The tetraspanin superfamily:

molecular facilitators.FASEB J.1997;11:428 – 442.

15. Satoh M, Ito A, Nojiri H, et al. Enhanced GM3 expression, associated with decreased invasiveness, is induced by brefel- din A in bladder cancer cells.Int J Urol.2001;19:723–731.

16. Miyake M, Nakano K, Itoi S-I, Koh T, Taki T. Motility-related protein-1 (MRP-1/CD9) reduction as a factor of poor prog- nosis in breast cancer.Cancer Res.1996;56:1244 –1249.

17. Huang C-I, Kohno N, Ogawa E, Adachi M, Taki T, Miyake M.

Correlation of reduction in MRP-1/CD9 and KA1/CD82 ex- pression with recurrences in breast cancer patients.Am J Pathol.1998;153:973–983.

18. Mori M, Mimori K, Shiraishi T, et al. Motility related protein 1 (MRP1/CD9) expression in colon cancer.Clin Cancer Res.

1998;4:1507–1510.

19. Sho M, Adachi M, Taki T, et al. Transmembrane 4 super- family as a prognostic factor in pancreatic cancer. Int J Cancer.1998;79:509 –516.

20. Higashiyama M, Doi O, Kodama K, et al. immunohisto- chemically detected expression of motility-related protein-1 (MRP-1/CD9) in lung adenocarcinoma and its relation to prognosis.Int J Cancer.1997;74:205–211.

TABLE 6

Multivariate Cox Proportional Hazards Regression Analysis of the Prediction of First Recurrence and First Progression among Patients with Pathologic Ta and T1 Tumors

Variable

Recurrence Progression

Crude

HR Pvalue 95% CI

Adjusted

HR Pvalue 95% CI

Crude

HR Pvalue 95% CI

Adjusted

HR Pvalue 95% CI

CD9 expression CD9 expression

Reduced vs. negative 0.81 0.481 0.44–1.47 0.94 0.846 0.51–1.74 0.55 0.103 0.27–1.13 0.69 0.335 0.33–1.47 Positive vs. negative 0.69 0.183 0.40–1.19 1.18 0.599 0.63–2.23 0.15 0.0001 0.08–0.31 0.28 0.007 0.11–0.70 Pathologic grade 1.45 0.001 1.17–1.79 1.61 0.003 1.18–2.20 2.80 0.0001 1.91–4.11 1.50 0.162 0.85–2.63 Pathologic stage 1.28 0.102 0.95–1.72 0.94 0.748 0.66–1.34 2.68 0.0001 1.58–4.54 1.40 0.288 0.75–2.61 Age 1.00 0.497 0.99–1.02 1.00 0.645 0.99–1.02 1.01 0.228 0.99–1.04 1.02 0.157 0.99–1.04 Gender 1.09 0.671 0.73–1.62 1.06 0.782 0.71–1.58 1.11 0.786 0.52–2.35 1.09 0.823 0.50–2.38 HR: hazard ratio; 95% CI: 95% confidence interval.

1656 CANCER October 15, 2003 / Volume 98 / Number 8

(10)

21. Uchida S, Shimada Y, Watanabe G, et al. Motility-related protein (MRP-1/CD9) and KAI1/CD82 expression inversely correlate with lymph node metastasis in esophageal squa- mous cell carcinoma.Br J Cancer.1999;79:1168 –1173.

22. Kusukawa J, Ryu F, Kameyama T, Mekada E. Reduced ex- pression of CD9 in oral squamous cell carcinoma: CD9 expression inversely related to high prevalence of lymph mode metastasis.J Oral Pathol Med.2001;30:73–79.

23. Mostofi FK, Sobin LH, Torloni H, editors. Histological typing of urinary bladder tumors. International classification of tumors, no. 10. Geneva: World Health Organization, 1973:

15–36.

24. Sobin LH, Wittekind C, editors. TNM classification of ma- lignant tumors, 6th ed. New York: John Wiley & Sons, 2002.

25. Smith JA, Labasky RF, Cockett ATK, Fracchia JA, Montie JE, Rowland RG. Bladder cancer clinical guidelines panel sum-

mary report on the management of nonmuscle invasive bladder cancer (stages Ta, T1 and TIS). J Urol.1999;162:

1697–1701.

26. Kroft SH, Oyasu R. Urinary bladder cancer: mechanisms of development and progression.Lab Invest.1994;71:158 –174.

27. Jordan AM, Weigarten J, Murphy WM. Transitional cell neoplasms of the urinary bladder. Can biologic potential be predicted from histologic grade? Cancer. 1987;60:

2766 –2774.

28. Holmang S, Hedelin H, Anderstrom C, et al. Prospective registration of all patients in a geographical region with newly diagnosed bladder carcinomas during a two-year pe- riod.Scand J Urol Nephrol.2000;34:95–101.

29. Bostwick DG, Ramnani D, Cheng L. Diagnosis and grading of bladder cancer and associated lesions.Urol Clin North Am.1999;26:493–507.

Références

Documents relatifs

at a time when this limitation was not yet in application and compulsory.” Although the formal/analytical grammar of the constitutional text seems to place it

Expression profiles of NGAL in peripheral blood mononuclear cells (PBMCs) from healthy individuals, untreated CLL patients and patient P16 before FCR treatment or at remission..

We used Iggy to study what are the regulatory events that explain the differential expres- sion between low and high aggressiveness from the KEGG interaction knowledge (network of

As we have stated, there is no general agreement about where the syllable boundaries lie in many words. Hence the word list was divided into three groups. The

Dès ce moment,, dans le beau temps comme dans la tempête, mal¬ gré les orages et les vents, malgré ceux qui ne voulaient pas le recon¬ naître, le chalet du régiment de

Table 1 shows the seven problem areas with the mean number of prob­ lems selected in each area for boys and girls.. The areas are ranked by number of problems chosen

Let us now, instead of viewing curriculum as something determined either locally or centrally, view it as something shaped by the broad policies of politicians

Le but de cette activité est de concevoir et d’utiliser un algorithme permettant de déterminer un encadrement de plus en plus fin du nombre √ 3..