Purpose: In this study, we investigated whether an oncofetal protein, IMP3, can serve as a new biomarker to predict progression and metastasis of early-stage urothelial carcinoma of the bladder.
Experimental Design: The expression of IMP3 in 242 patients with primary superficial bladder urothelial carcinoma and metastatic urothelial carcinoma was evaluated by immunohistochemistry. Patients with primary superficial urothelial carcinoma of the bladder were further investigated by use of survival analysis.
Results: Twenty percent (42 of 214) of primary superficial urothelial carcinomas and 93% (26 of 28) of metastatic urothelial carcinomas expressed IMP3. Kaplan-Meier plots and log-rank tests showed that patients with IMP3-positive tumors had a much lower progression-free survival (P = 0.0002) and disease-free survival rate (P = 0.0067) than did those with IMP3-negative tumors. The 5-year progression-free and disease-free survival rates were 91% and 94% in IMP3-negative patients versus 64% and 76% in IMP3-positive patients, respectively. Sixty percent of IMP3-positive patients with superficial invasive urothelial carcinoma at initial diagnosis went on to develop metastases, whereas no metastasis was found in IMP3-negative patients (P = 0.0017). In the multivariable Cox analysis, patients with IMP3 expression in their superficial urothelial carcinomas subsequently developed invasive tumors or metastasis at a rate that was about five times greater than cases without expression of IMP3 adjusting for other well-known clinical variables (tumor stage and grade, etc.).
Conclusions: Our findings indicate that IMP3 is an independent prognostic marker that can identify a group of patients with a high potential to develop progression and who might benefit from early aggressive therapy.
- superficial urothelial carcinoma
- prognostic biomarker
- oncofetal protein
- Cell Growth/Signaling Pathways
- Genitourinary cancers: bladder
- Protease-inhibitor systems
- Tumor promotion and progression
Urothelial carcinoma is the most common type of urinary bladder cancer and accounts for about 95% of malignant bladder tumors. Each year, nearly 63,000 new cases of urothelial cancer are diagnosed with almost 13,000 people dying of the disease (1, 2). Seventy-five percent of urothelial carcinomas of the bladder are superficial tumors including the tumor, node, and metastasis (TNM) categories of Ta (papillary tumors confined to the epithelium), Tis (flat carcinoma in situ [CIS]) also confined to the epithelium, and T1 (tumors invading the lamina propria; refs. 2, 3). The biological behavior of these superficial urothelial cancers is significantly different, ranging from the relatively benign noninvasive papillary tumor to the highly aggressive tumor with a significant mortality rate (2). Currently, identification of patients with aggressive superficial urothelial carcinoma and determination of their treatment are mainly based on the tumor grade and stage (3). Treatment for the majority of low-grade and Ta superficial urothelial carcinoma is local resection of the tumor with close observation or intravesical therapy, whereas more aggressive therapies including cystectomy and/or radiation/chemotherapy are often reserved for patients with high-grade or T1 tumors who have a greater chance to progress to deeply invasive cancer with a much higher mortality rate (3–8). However, the tumor grade and stage have limited ability to predict tumor progression. As superficial urothelial carcinomas show unpredictable behavior, an important clinical problem is how to accurately assess individual risk of progression and to stratify patients for treatment. Therefore, there is a great need for biomarkers that can accurately distinguish superficial tumors with a high probability of progression from those that will remain indolent. Using such biomarkers, one could predict the patient's prognosis and effectively target the individuals who would most likely benefit from the therapy.
Recently, we identified IMP3 as an independent prognostic biomarker with an expression that predicts aggressive behavior of renal cell carcinomas (9). IMP3 is a member of the insulin-like growth factor II mRNA binding protein (IMP) family that consists of IMP1, IMP2, and IMP3 (10). IMP family members play an important role in RNA trafficking and stabilization, cell growth, and cell migration during the early stages of embryogenesis (11). The IMP3 gene is located on chromosome 7p11.2 (12) and is identical to the KOC (KH domain containing protein overexpressed in cancer) protein that was originally cloned from a pancreatic tumor cDNA screen (13). IMP3 is expressed in developing epithelium, muscle, and placenta during early stages of human and mouse embryogenesis, but it is expressed at low or undetectable levels in adult tissues (10, 11). The expression of IMP3/KOC is also found in malignant tumors including pancreas, lung, stomach, and colon cancers as well as soft tissue sarcomas, but it is not detected in adjacent benign tissues (10, 13–15). Moreover, a recent study has shown that IMP3 promotes tumor cell proliferation and invasion (16, 17). These findings indicate that IMP3 is an oncofetal protein that may have a critical role in the regulation of cell proliferation and invasion. In this study, we investigated whether IMP3 could serve as an independent biomarker to predict progression and metastasis of superficial urothelial carcinoma.
Materials and Methods
Patients and tumor specimens. Formalin-fixed, paraffin-embedded samples of bladder biopsies from 214 patients with a diagnosis of urothelial carcinoma of the bladder (Ta [noninvasive papillary carcinoma], n = 171; Tis [CIS], n = 14; T1[tumor invades lamina propria without evidence of muscularis propria invasion], n = 29) were obtained at the University of Massachusetts Memorial Medical Center in Worcester MA, from 1992 to 2000. All T1 tumors were confirmed by tissue diagnosis. These biopsies showed the presence of lamina propria with no tumor presence in muscularis propria. The cases with lamina propria invasion but without the presence of muscularis propria were excluded to rule out a potential higher stage of urothelial tumors. The cases with T1 diagnosis in original biopsy but rebiopsy within 3 months showing tumor invasion in muscularis propria (>T1 tumor) were also excluded in our study. An additional 28 metastatic urothelial bladder carcinoma specimens were included in the study and evaluated for the expression of the IMP3. Tumor grade and stage were based on pathologic findings following the World Health Organization grading system (World Health Organization 2004) and the American Joint Committee on Cancer staging manual, sixth edition, respectively. The data from these sources represented all patients for whom archival tissues and adequate clinical follow-up information were readily available. The Institutional Review Board approved this study.
Immunohistochemical analysis. Immunohistochemical studies were performed on 5-μm sections of formalin-fixed, paraffin-embedded tissue from bladder biopsy specimens as a previously published protocol (15). Antigen retrieval was carried out with 0.01 mol/L citrate buffer at pH 6.0, in an 800-W microwave oven for 15 minutes before immunostaining. The slides were stained on the DAKO Autostainer (DAKO Corporation, Carpinteria, CA) using the EnVision (DAKO) staining reagents. The sections were first blocked for endogenous protein binding and peroxidase activity with an application of Dual Endogenous Block (DAKO) for 10 minutes, followed by a buffer wash. The sections were then incubated with a mouse monoclonal antibody specific for IMP3 (L523S, DAKO) at a 2.0 μg /mL concentration for 30 minutes, followed again by a buffer wash. Sections were then incubated with the EnVision+ Dual Link reagent (a polymer conjugated with goat anti-mouse Ig, and horseradish peroxidase) for 30 minutes. The sections were then washed and treated with diaminobenzidine (DAB) and hydrogen peroxide, to produce the visible brown pigment. A toning solution (DAB Enhancer, DAKO) was used to enrich the final color. The sections were counterstained with hematoxylin, dehydrated, and coverslipped with permanent media.
Sections of pancreatic carcinoma with known positivity of IMP3 were used as positive controls for the L523S mouse monoclonal antibody (MAb) specific for IMP3/KOC (Corixa Corporation, Seattle, WA) staining. Negative controls were performed by replacing the primary antibody with nonimmune IgG. Positive staining of IMP3 was defined as a dark-brown cytoplasm staining pattern in the tumor epithelial cells, which can be easily observed at low-power magnification (<40×). Scant fine granular background staining of epithelial cells, which cannot be seen at low-power magnification (≤40×), or no staining at all was considered negative. The status of IMP3 was assessed by researchers LS and ZJ without knowledge of the clinical and pathologic features of the cases or the clinical outcome.
Statistical analysis. Tumor progression was measured as the date when tumors were shown on biopsy to have evolved to a higher stage (deeper invasion) or when metastases were proven by clinical or pathologic diagnosis. No progression was defined as negative for tumor or no change from the original T stage in a follow-up biopsy and/or negative cystoscopic findings with at least two negative cytology results at the last follow-up date. Recurrence was defined as the appearance of recurrent urothelial carcinoma without alteration from the original stage. Progression-free survival was measured from the date of the biopsy to the date of follow-up tissue biopsy/resection that showed tumor progression to a higher stage of tumor as compared with the previous biopsy, or the date of the last negative cystoscopy with negative cytology, or the last tissue diagnosis proved without progression. Metastasis-free survival was measured from the date of the biopsy to the date of first clinical evidence of metastasis and was censored at the date of death or last follow-up visit for survivors. The patients' disease-free survival was measured from the date of the diagnostic biopsy to the date of clinical evidence of remaining urothelial carcinoma and was censored at the date of death without disease or last follow-up for survivors. Overall survival was measured from the date of biopsy to the date of death and was censored from the date of the last follow-up for survivors. Data on various treatments including transurethral resection of the tumor in the bladder, BCG treatment, intravesical therapy, radiation and chemotherapy, and cystectomy were collected during the time of each follow-up category. The patients' age, gender, grade, stage, tumor size, multiplicity, recurrence, and various treatment modalities were collected as baseline variables. The distribution of each variable was compared for IMP3-positive and IMP3-negative subgroups with the Wilcoxon rank-sum test for continuous variables and the Fisher's exact test for categorical variables. Progression-free survival, metastasis-free survival, disease-free survival, and overall survival were estimated by Kaplan-Meier method and evaluated with the use of log-rank test for univariate analysis. The Cox proportional hazards model was used to assess simultaneous contribution of baseline covariates in univariate and multivariable analysis. A two-sided P value of less than 0.05 was considered to be statistically significant.
Expression of IMP3 in all primary and metastatic urothelial carcinomas. When detected by IHC, IMP3 protein was observed in the cytoplasm of tumor cells yielding a dark-brown staining pattern (Fig. 1A ). In contrast, cases that were scored as having no expression of IMP3 completely lacked the brown stain (no staining, Fig. 1B). No IMP3 expression was found in benign tissue adjacent to cancer. Expression of IMP3 was found in 20% (42 of 214) of primary superficial urothelial carcinomas, whereas 93% (26 of 28) of metastatic urothelial carcinomas were positive for IMP3 staining.
Patient characteristics.Table 1 shows the relevant clinical characteristics of the 214 patients with superficial urothelial carcinoma of the bladder. The expression of IMP3 was strongly associated with tumor stage (P = 0.016), grade (P < 0.0001), recurrent rate (P = 0.029), and treatment (P = 0.001), and the patient's age (P = 0.903), gender (P = 0.497), tumor size (P = 0.237), and multiplicity (P = 0.692) were not linked with IMP3 positivity. The IMP3 positivity rate was increased in T1 (35%) and urothelial CIS (Tis, 36%) as compared with Ta tumors (16%, P = 0.016). No papillary urothelial neoplasm of low malignant potential expressed IMP3, whereas 18% to 42% of low- and high-grade urothelial carcinomas were positive for IMP3 (P < 0.0001).
Expression of IMP3 and tumor prognosis. The progression-free survival rate was significantly different between patients with IMP3-positive tumors versus those that were IMP3 negative. By the end of the follow-up, 26% (11 of 42) of patients with IMP3 positivity in their superficial urothelial cancers subsequently developed either deeper invasion or metastasis (median follow-up = 35 months, range = 3 to 125 months), whereas only 7% (12 of 172) of patients without expression of IMP3 were found to show progression (median follow-up = 45 months, range = 2 to 146 months). Twenty-two patients whose status (progression versus non progression) was unknown (no follow-up tissue biopsy) were all negative for IMP3 in their original tumors. In Ta disease, 19% (5 of 27) of IMP3-positive patients (median follow up = 38 months, range = 6 to 125 months) versus 5% (7 of 144) of IMP3-negative patients (median follow-up = 47 months, range = 2 to 146 months) showed progression (P = 0.0098). In T1 disease, 60% (6 of 10) of IMP3-positive patients (median follow-up = 15 months, range = 3 to 72 months) versus 21% (4 of 19) of IMP3-negative patients (median follow-up = 17 months, range = 2 to 88 months) were found to have tumor progression (P = 0.03).
Kaplan-Meier plots and log-rank tests in patients with superficial urothelial cancer showed that patients without IMP3 expression in their superficial urothelial carcinomas had significant longer progression-free survival (P = 0.0002; Fig. 2A ) and disease-free survival (P = 0.0067; Fig. 2B) than patients with IMP3 expression. The 5-year progression-free survival rate was 91% in IMP3-negative patients versus 64% in IMP3-positive patients. The 5-year disease-free survival rate was 94% in patients without expression of IMP3 versus 76% in patients with IMP3 expression. In patients with superficial urothelial cancers at stage Ta and T1, the status of IMP3 expression was significantly associated with an increased risk of progression (Ta: Fig. 3A , P = 0.0098; T1: Fig. 3B, P = 0.03). Metastasis was not found in patients with Ta and Tis disease.
Figure 4 shows that IMP3-positive patients with T1 disease had significant poorer metastasis-free survival compared with IMP3-negative patients with T1 disease (P = 0.0017). There were no significant differences in overall survival (P = 0.39) between patients with IMP3-positive staining and patients with IMP3-negative staining. No significant difference in disease-free survival was found between IMP3-positive and IMP3-negative patients with T1 disease (P = 0.20).
Multivariate analysis. The results of the multivariate analysis for progression-free survival and disease-free survival in the 214 patients with superficial urothelial cancers are presented in Table 2 . For these analyses, risk factors shown in Table 1 were initially included in the model as potential risk factors. Multivariate Cox proportional hazards regression analysis showed that the expression of IMP3 in superficial urothelial cancers was a strong independent predictor of the patients' clinical outcome. The hazard ratio for progression-free survival was 6.46 (95% confidence interval, 2.19 to 19.05; P = 0.001) for the patients with IMP3 expression. The hazard ratio for disease-free survival was 2.82 (95% confidence interval, 1.18 to 6.71; P = 0.019) for the patients with IMP3 expression. In addition to IMP3 status, age and tumor stage T1 (compared with stage Ta) were also observed as significant risk factors for progression-free survival, and age, tumor stage T1 (compared to stage Ta), and BCG treatment were also observed as significant factors for disease-free survival (Table 2).
Due to the limited number of patients, the risk factor of coexistent CIS (N = 7) and patients' treatment including radiation therapy (N = 5), cystectomy (N = 9), and systematic chemotherapy (N = 4) were not included in the analysis.
Different cases of morphologically identical superficial bladder cancer can behave very differently. In this study, we have shown that IMP3 is expressed much more frequently in metastatic urothelial carcinoma than in primary urothelial carcinoma and that IMP3 expression is associated with tumor progression. Therefore, IMP3 is a biomarker that can provide important prognostic information in patients with superficial urothelial cancers.
Many tumor markers have been studied for their potential use in assessing the prognosis of urothelial carcinoma. However, a recent international consensus panel on prognostic markers for bladder cancer concluded that although certain markers such as p53 appear to be promising in predicting progression of bladder cancer, the data were still heterogeneous (18). None of the markers have been adopted in clinical practice (18). Although markers such as p53 have been extensively studied and some of them have shown promising results, a recent meta-analysis and review indicates that evidence is not sufficient to conclude whether alterations in p53 expression act as markers of outcome in patients with bladder cancer (19). Thus, we still lack prognostic biomarkers for urothelial carcinoma.
IMP3 as a prognostic marker exhibits several attractive features for superficial urothelial carcinoma. First, the expression of IMP3 is correlated with other known aggressive indicators for superficial urothelial carcinoma. Our results showed that the expression of IMP3 was strongly related to higher tumor grade and stage and tumor recurrence. Second, the expression of IMP3 in biopsies of superficial urothelial carcinomas is an independent predictor of tumor progression. We found significantly increased tumor progression in patients with superficial urothelial carcinomas expressing IMP3 as compared with those without IMP3 expression. These findings were independent of tumor grade and stage. In the multivariable Cox analysis, patients with IMP3 expression in their superficial urothelial carcinomas subsequently developed invasive lesions or metastasis at a rate more than six times greater than cases without IMP3 expression adjusting for other well-known clinical variables. Therefore, IMP3 status in the initial tumor biopsies is a potentially important new risk factor that may be able to be used in addition to tumor stage, grade, size, multiplicity, and coexistent CIS to guide the decision of adjuvant courses of intravesical therapy after tumor resection. Third, IMP3 expression is associated with metastasis of urothelial carcinoma. Our data showed that the expression of IMP3 was significantly increased not only in metastatic urothelial carcinoma but most importantly also in patients with primary T1 tumors who subsequently developed metastatic disease. We found that 60% of patients with IMP3 positivity in their primary T1 urothelial carcinomas urothelial carcinomas developed metastasis, whereas none of the patients without IMP3 expression in their primary T1 tumors developed metastases. Currently, the decision on whether to attempt bladder conservation with intravesical therapy or to perform a cystectomy is the most difficult issue in managing superficial bladder cancer. Results from early cystectomy for high-risk superficial urothelial carcinomas are generally excellent with 5-year cancer-specific survival in exceeding 90% (6, 7). However, as patients who undergo cystectomy have a significantly unpleasant lifestyle and in the absence of better prognostic tools, many patients who would not have progressed are subjected to these potential side effects (3, 20, 21). Therefore, accurately identifying the high-risk patients, particularly those who have T1 tumors with poor prognosis, becomes a very important clinical issue. The ability of IMP3 to identify patients presenting at stage T1 that will progress and present with metastatic disease would provide very important clinical information. Presumably, if this subset of patients could be identified prospectively (with IMP3), they would benefit from earlier definitive treatment (cystectomy), and this concept could be tested in future clinical trials.
Fourth, IMP3 immunohistochemical staining is a simple, inexpensive and reliable assay, which can be used in routine clinical practice. As superficial urothelial carcinomas are usually treated by resection, tumor tissue is routinely available for immunohistochemical staining. There are also two important factors to reduce inconsistent or variable results of IHC: (1) The IMP3 monoclonal antibody specifically binds its target with very low or no background in immunohistochemical staining and (2) The utilization of a standard protocol and automated instrument minimizes variations in the staining results.
Although there are significant differences in disease progression and disease-free survival rates between patients with IMP3-positive staining and patients with IMP3-negative staining, we did not find significant differences in overall survival rates. One of the potential reasons for this difference is that most patients with superficial urothelial carcinomas have a long survival and may die from diseases other than bladder cancer. We did not find a significant difference in disease-free survival between IMP3-positive and IMP3-negative patients with T1 disease, presumably because of the limited number of patients. However, our results provide a very strong rationale for initiating a much larger clinical study, particularly for patients with T1 disease, to further validate this marker.
Our findings raise the possibility that IMP3 may play a direct role in the progression and metastasis of urothelial carcinoma. There are emerging data that IMP3 may play a role in the growth of malignant cells and cellular de-adhesion (3, 20, 21). Interestingly, Yaniv et al found that IMP3 in Xenopus laevis is required for the migration of cells forming the roof plate of the neural tube and, subsequently, for neural crest migration (22), which suggested that IMP3 may play an important role in promoting cell migration. These findings could explain why IMP3 is associated with tumor progression and metastasis. Further study is required to investigate whether IMP3 plays a direct role in the biological behavior of urothelial carcinoma.
In summary, our findings provide evidence that IMP3 is an independent prognostic marker that can be used at the time of initial diagnosis of superficial urothelial carcinomas to identify a group of patients with a high potential to develop progression and metastasis and those who might benefit from early aggressive therapy.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Received August 20, 2007.
- Revision received October 24, 2007.
- Accepted November 8, 2007.