|
Advertisement |
|
|
||
 |
|
|||||
|
|||||
|
 |
||||||
|
||||||
|
||||||
|
© 2002 American Society for Clinical Oncology DPC4/Smad4 Expression and Outcome in Pancreatic Ductal AdenocarcinomaByFrom the Cancer Research Program, Garvan Institute of Medical Research, and Division of Surgery and Department of Anatomical Pathology, St Vincent’s Hospital Campus, Darlinghurst, Sydney; Department of Anatomical Pathology, Royal Prince Alfred Hospital, and Department of Pathology, University of Sydney, Camperdown; and Institute of Clinical Pathology and Medical Research, Westmead Hospital, Westmead, New South Wales, Australia. Address reprint requests to S.M. Henshall, PhD, Cancer Research Program, Garvan Institute of Medical Research, 384 Victoria St, Darlinghurst, Sydney, New South Wales 2010, Australia; email: s.henshall{at}garvan.org.au
PURPOSE: Prognostic indicators in pancreatic cancer (PC) are poorly defined and difficult to quantify preoperatively, hence they may lead to inappropriate patient selection for treatment. We examined the protein expression of key cell-cycle regulatory and cell-signaling molecules that occur at high frequency in PC and assessed their relationship to clinicopathologic parameters, response to operative resection, and outcome. PATIENTS AND METHODS: We identified 348 patients with pancreatic ductal adenocarcinoma and assessed the influence of reported clinicopathologic prognostic factors and the expression of the cell-cycle regulatory genes p21WAF1/CIP1 (CDKN1A), cyclin D1 (CCND1), p53, and p16INK4A (CDKN2) and the cell-signaling molecule DPC4/Smad4 (MADH4) using immunohistochemistry in a subgroup of 129 patients. RESULTS: Independent prognostic factors in resected patients were tumor size greater than 45 mm (P = .0015), involvement of surgical margins (P < .0001), and perineural invasion (P = .014). Loss of DPC4/Smad4 expression cosegregated with resectability (P < .0001) and was associated with improved survival after resection (P < .0001), whereas resection did not improve survival in patients whose tumor expressed DPC4/Smad4 (P = .5). Aberrant expression of p21WAF1/CIP1, cyclin D1, p53, or p16INK4A was not associated with a difference in survival. CONCLUSION: Tumor size (> 45 mm), resection margin involvement, and perineural invasion were independent prognostic factors. Preoperative assessment of DPC4/Smad4 expression has potential as a prognostic indicator in patients with PC since resection did not benefit those patients whose cancers expressed DPC4/Smad4 and accurate assessment of DPC4/Smad4 expression, unlike tumor size, margin status, and perineural invasion, does not require resection.
DUCTAL ADENOCARCINOMA of the pancreas is the fifth leading cause of cancer-related death in Western societies.1 Pancreatic cancer (PC) presents as clinically advanced disease. As a result, the mortality from this disease almost parallels its incidence, with a 5-year survival rate of less than 10%.2 The majority of patients who present with PC do not undergo operative treatment, and nonoperative treatments offer little, if any, survival advantage. There are few substantial series reporting significant prognostic markers in PC. The influence of factors such as tumor size, lymph node involvement, status of resection margins, DNA ploidy, degree of differentiation, and perineural invasion are inconclusive.3-6 Inconsistencies in the influence of clinicopathologic factors on outcome in PC suggest that some of these parameters may be inaccurate predictors of prognosis and response to treatment. In addition, preoperative estimation of tumor size and lymph node involvement is difficult. As a consequence, patient selection for operative intervention based on preoperative estimation of these parameters may be inappropriate. In other cancers, gene mutations, polymorphisms, and aberrations in gene expression have cosegregated with therapeutic response and prognostic groups, eg, overexpression of estrogen receptors and HER2/neu in breast cancer. Hence, molecular markers may assume greater importance in estimating prognosis and selection for treatment for patients with PC. They may be either independently prognostic or surrogate markers of other prognostic markers, such as lymph node involvement, that are difficult to determine without resection. Dysregulation of the normal cell-cycle regulatory machinery is integral to the neoplastic process, and there is now compelling evidence implicating loss of cell-cycle control in the development and progression of most human cancers.7 Abnormalities in the retinoblastoma pathway that controls G1- to S-phase progression in the cell cycle and the transforming growth factor beta (TGF-ß) signaling pathway occur at high frequency in pancreatic ductal adenocarcinomas.8 Mutations of the tumor suppressor gene p53 occur in approximately half of all human malignancies,9 and the inactivation of p53 has been consistently reported in 50% to 75% of PC cases.1,10,11 p53 protein has many biologic functions, including the regulation of programmed cell death. In relation to cell-cycle control, it inhibits cell proliferation by increasing intracellular levels of the cyclin dependent kinase (cdk) inhibitor p21WAF1/CIP1.9,12 p21WAF1/CIP1 prevents phosphorylation of the retinoblastoma gene product (pRb) by binding to the cyclin D1/cdk4 and cyclin E/cdk2 enzyme complexes, thereby preventing formation of the active enzymes that are required for pRb phosphorylation.7 Cyclin D1 overexpression has been identified as a marker of poor outcome in several human cancers13 and occurs in 68% of PCs, where it was also associated with a poor prognosis.14 The tumor suppressor gene INK4A maps to chromosome 9p21 and encodes the cdk inhibitor p16INK4A, a key regulator of cell-cycle progression at the G1–S-phase transition.15 p16INK4A binds to cdk4 and cdk6 to inhibit their catalytic activity, which leads to reduced phosphorylation of pRb and hence G1 cell-cycle arrest.15,16 p16INK4A is frequently inactivated in human cancer,15 and in pancreatic cancer, p16INK4A is inactivated in more than 90% of ductal adenocarcinomas.1,17 The mechanisms of p16INK4A inactivation include homozygous deletion, mutation, and promoter methylation.18 The TGF-ß growth inhibitory pathway utilizes Smad proteins 1 through 8. Characterization of Smad functions has segregated them into three groups: receptor-regulated Smads (Smads 1, 2, 3, 5, and 8), common partner Smad (Smad4), and inhibitory Smads (Smads 6 and 7).19 Normally, ligand-induced TGF-ß receptor activation results in the formation of heterodimeric complexes of Smad3 with Smad4 leading to the translocation of these complexes into the nucleus, where Smad4 activates transcription of cell-cycle inhibitory factors p21WAF1/CIP1 20,21 and p15INK4B.22 Homozygous deletion or mutation of DPC4, the gene encoding Smad4 located on chromosome 18q, has been reported in 55% of pancreatic ductal adenocarcinomas.8 Loss of Smad4 may lead to upregulation of the Rb pathway, consequent progression from the G1 to the S phase of the cell cycle, and hence increased cellular proliferation. We assessed the influence of clinicopathologic parameters on prognosis in a cohort of 348 patients with PC accrued retrospectively from three teaching hospitals in Sydney, Australia. In addition, we assessed the influence of molecular aberrations that occur at high frequency in PC (p21WAF1/CIP1, cyclin D1, p53, p16INK4A, and DPC4/Smad4) on patient outcome in a subgroup of 129 patients with long-term follow-up treated for pancreatic ductal adenocarcinoma.
We identified a cohort of 348 patients with the diagnosis of pancreatic ductal adenocarcinoma from Westmead Hospital, St Vincent’s Hospital Campus, and Royal Prince Alfred Hospital in Sydney, Australia. Archival formalin-fixed, paraffin-embedded tissue from 129 pancreata that were resected or biopsied between January 1972 and July 1999 was available. Multicenter ethical approval for data collection and tissue use was granted by the St Vincent’s Hospital, Westmead Hospital, and Royal Prince Alfred Hospital ethics committees.
Clinicopathologic Parameters
Immunohistochemistry
Immunohistochemical Scoring
Statistical Analysis
Cohort Characteristics The cohort (Table 1) consisted of 348 patients with the diagnosis of PC (190 men and 158 women). The average age at diagnosis was 64 years (median age, 67 years; range, 34 to 94 years). Of 129 patients for whom tissue was available, 51 samples were from pancreatic resections (48 Whipple pancreaticoduodenectomies and three distal pancreatectomies), 74 samples were from intraoperative incision biopsies, and four samples were postmortem specimens that were included in analysis as part of the natural history of the disease. For the remaining patients within the cohort, a clinical diagnosis of PC was made, 11 with supportive fine-needle aspiration biopsy cytology results. PCs arising from intraductal papillary mucinous tumors or mucinous cystic neoplasms were not included in this cohort. Median follow-up for the cohort was 3.5 months (range, 0 to 117 months). Only five patients were alive at the census date (August 21, 2000). Of the 348 patients, 319 (92%) died of PC and 14 (4%) died of other causes (10 were lost to follow-up). The overall median survival time was 3.4 months; the disease-specific survival time was 3.5 months. The overall disease-specific 1-year survival rate was 13%, with a 5-year survival rate of 3.3%. Of 129 tumors with available tissue, 113 (88%) were of the tubular type; the remaining 16 (12%) were of the mucinous noncystic or colloid type. The majority of tumors were either moderately differentiated (n = 64) or poorly differentiated (n = 51); only 14 were well differentiated. Twenty-three tumors were categorized as International Union Against Cancer32 stage I, six as stage II, 76 as stage III, and 24 as stage IV (Table 1). Adjuvant therapy for PC has not been routinely offered in the hospitals involved in the study. We identified 14 (11%) of 129 patients who received any form of chemotherapy or radiotherapy that was commenced within 3 months of their operation. Of these patients, eight had undergone tumor resection and six had undergone operative biopsies only. Protocols for chemotherapy and radiotherapy were not standardized, but chemotherapy was fluorouracil-based. Adjuvant therapy did not influence survival (any "adjuvant" therapy: overall, logrank P = .7498, resected, logrank P = .8992, biopsied only, logrank P = .5803; postoperative chemoradiation: resected, logrank P = .7102).
On Kaplan-Meier analysis, operative resection of the tumor was associated with longer survival than either operative biopsy or no operative intervention (median survival, 10.4, 4.7, and 2.1 months, respectively; logrank, P < .0001; Fig 1A). Patients with stage I and II tumors (lymph node–negative) survived significantly longer (median survival, 16.2 months) than those with stage III and IV tumors (median survival, 6.4 months) (logrank, P < .0001; Fig 1B). Those with poorly differentiated tumors had a shorter survival (median, 4.7 months) than those with well or moderately differentiated tumors (median, 8.5 months) (logrank, P = .0044; Fig 1C). On multivariate analysis using the Cox proportional hazards model for those factors that demonstrated a significant effect on survival on univariate analysis, resection and stage were the only independent prognostic factors when modeled together with degree of differentiation and DPC4/Smad4 status (Table 2, A and B). Subsequent analysis demonstrated that although differentiation was independent of operative treatment, it was not independent of stage (Table 2, D and E). Well-differentiated and moderately differentiated tumors were associated with resectability (  2, P = .0083) and stage (2, P = .0184), which likely accounts for why differentiation was not an independent prognostic factor on multivariate analysis. Operative treatment was also associated with stage (2, P < .0001)
Loss of DPC4/Smad4 expression occurred in 63 (53%) of 119 tumors (Table 1). When DPC4/Smad4 expression was present within the tumor, more than 95% of cells exhibited cytoplasmic staining, with more than 80% also exhibiting nuclear staining. Since the cohort spanned 25 years, staining was compared between specimens obtained before and after 1990. There was no difference in the staining rates in specimens obtained before or after 1990 (P = .7902). In DPC4/Smad4-negative specimens, occasional focal staining of fewer than 10 cells was seen, so a cutoff of 5% was used. Focal positive staining likely represents a nondominant clone of the tumor.33 Detectable expression of DPC4/Smad4 was associated with a poor outcome on univariate analysis (logrank, P = .0088) (Fig 1D). On multivariate analysis, DPC4/Smad4 expression was not an independent prognostic indicator (Table 2, A). Loss of DPC4/Smad4 expression was associated with resectability, as 35 (58%) of 60 DPC4/Smad4-negative tumors were resected compared with 10 (17%) of 59 DPC4/Smad4-positive tumors ( 2, P < .0001). This suggests confounding of these two variables to explain the difference in survival seen on univariate analysis. DPC4/Smad4 status was only independent of the degree of differentiation; it was not independent of operative treatment or stage (Table 2, F, G, and H). Loss of DPC4/Smad4 expression was also associated with early-stage tumors (2, P = .0005) but not with degree of differentiation (2, P = .7926). There was no difference in patient outcome among the three treatment centers (P = .1058). The year of treatment only influenced outcome when those patients treated before 1990 were compared with those treated from 1990 onward (P = .0018; Table 1). This was due to a nonstatistically significant difference in the resection rate between these time periods and hence was not an independent prognostic factor (Table 2, I). Most importantly, operative resection did not benefit those patients whose tumors expressed DPC4/Smad4 (logrank, P = .5093; Fig 2A), but it was beneficial in those who had lost DPC4/Smad4 expression (logrank, P < .0001; Fig 2B). Furthermore, survival for patients with resected DPC4/Smad4-negative tumors was significantly longer than survival for all other groups combined (Fig 2C). Hence, loss of DPC4/Smad4 expression was associated with resectability of PC in this cohort, with only those patients who had loss of DPC4/Smad4 expression benefiting from the resection.
p53 was overexpressed in 67 (54%) of 125 patients, cyclin D1 was overexpressed in 60 (50%) of 121 patients, loss of p16INK4A expression occurred in 68 (69%) of 98 patients, and p21WAF1/CIP1 was overexpressed in 99 (79%) of 125 patients (Table 1, Fig 3). Neither p53, cyclin D1, p16INK4A, nor p21WAF1/CIP1 expression status cosegregated with patient outcome or with each other.
Because DPC4/Smad4 exerts its cell-cycle inhibitory effects in part through increasing transcription of p21WAF1/CIP1, we examined the relationships of these two molecules to each other and to clinicopathologic parameters. In our cohort of 59 tumors that demonstrated loss of DPC4/Smad4 expression, 42 (71%) still showed p21WAF1/CIP1 overexpression, and of these 42 tumors, 33 (79%) were resected compared with only two (12%) of 17 tumors that demonstrated loss of DPC4/Smad4 expression but had low p21WAF1/CIP1 expression (P < .0001; Table 3). Hence, a proportionally greater number of patients whose tumors demonstrated low p21WAF1/CIP1 expression (within the DPC4/Smad4-negative subgroup) only had biopsies and did not proceed to resection. On Kaplan-Meier analysis for all patients, those with loss of DPC4/Smad4 expression and p21WAF1/CIP1 overexpression had a favorable prognosis (median survival, 9.2 v 5.2 months, n = 113; logrank, P = .0021; Fig 2D). On multivariate analysis (data not shown), patients with p21WAF1/CIP1 overexpression in the setting of loss of DPC4/Smad4 expression did not differ from those patients with tumors that demonstrated loss of DPC4/Smad4 expression alone (Table 2). In addition, in the subgroup of patients whose tumors demonstrated loss of DPC4/Smad4 expression, p21WAF1/CIP1 overexpression showed a trend toward improved survival (logrank, P = .0594).
Resected Pancreatic Cancer Of the 129 patients who underwent laparotomy (Table 4), 51 proceeded to pancreatic resection (28 male patients; median age, 61 years; range, 34 to 83 years). Median follow-up was 9.5 months (range, 0 to 117 months) with a median disease-specific survival time of 10.4 months, 1-year survival rate of 41%, and 5-year survival rate of 11%. The only patients still living in the cohort (n = 5) underwent resection. The majority (40 patients) died of PC, two died of non-PC causes, and two were lost to follow-up. The 30-day mortality for resection was 4% (two patients). Twenty-seven (53%) of the resected pancreata had lymph nodes free of tumor. The mean tumor size was 37 mm (range, 15 to 75 mm). Resection margins were microscopically free of tumor in 28 cases (55%). Perineural invasion, defined as tumor present within the nerve sheath, not contiguous or adjacent to infiltrating tumor microscopically, was present in 38 (80%) of the resected specimens. Poorly differentiated tumors occurred in 13 patients (25%), with five well-differentiated (10%) and 33 moderately differentiated (65%) tumors (Table 4).
On Kaplan-Meier analysis, those patients whose resected pancreata showed margins clear of microscopic disease had a survival advantage (median survival, 16.2 months) compared with patients with tumor at the resection margin (median survival, 6.2 months) (P < .0001), as did those who had tumors  45 mm (median survival, 13.6 v 6.0 months; P = .0015). Patients with involvement of their pancreatic resection margins by tumor (median survival, 6.2 months) had no survival advantage over those who had laparotomy with biopsy only (median survival, 4.7 months; P = .4177). Patients whose resected pancreata showed no tumor involvement of lymph nodes survived longer (median survival, 16.2 v 8.6 months; P = .0137), as did those who did not exhibit perineural invasion (median survival, 12.4 v 9.2 months; P = .0140). Loss of DPC4/Smad4 expression was observed in 35 (78%) of 45 resected pancreata. Its relationship to other parameters is shown in Table 5. Loss of DPC4/Smad4 expression was associated with a favorable prognosis (median survival, 13.6 v 6.4 months; logrank, P = .0257) (Fig 4E). The center where the patients underwent resection had no influence on outcome (P = .1627). The majority of patients (88%) were treated in the 1990s. The only difference in outcome by year of treatment was when those treated before 1990 were compared with those treated during or after 1990 (P = .0156). Results of multivariate analyses of those parameters found to be prognostic on univariate analysis are shown in Table 6. Tumor size, margin status, and perineural invasion were all independent prognostic factors (Table 6, A). DPC4/Smad4 expression was only of borderline significance, most likely as a result of its relationship to tumor size (Table 6, B). Lymph node status and DPC4/Smad4 expression were confounding variables (Table 6, C and D), with DPC4/Smad4 expression being a more important prognostic factor than lymph node status (Table 6, E) because lymph node status was more strongly related to tumor size than DPC4/Smad4 expression (Table 6, B and F). Tumor size had a major effect on other parameters in the multivariate model, and when excluded, DPC4/Smad4 expression was an independent prognostic factor (Table 6, G), whereas lymph node status did not maintain statistical significance (Table 6, H). Although inclusion of year of treatment in the model showed that loss of DPC4/Smad4 expression was independently prognostic (Table 6, I), there were only six resections in the pre-1990 group, and hence the data are unlikely to be reliable when six parameters are modeled together.
Univariate analysis showed that patients whose resected pancreata demonstrated loss of DPC4/Smad4 expression and p21WAF1/CIP1 overexpression, as demonstrated in the overall group above, was of borderline significance (logrank, P = .0494) and was not a better indicator than loss of DPC4/Smad4 alone (logrank, P = .0257), with similar results on multivariate analysis (data not shown). Only two of 35 resected patients who demonstrated loss of DPC4/Smad4 expression had low p21WAF1/CIP1 expression, precluding survival analysis in subgroups based on DPC4/Smad4 status, as was performed for the full cohort.
Biopsied Pancreatic Cancer
The size of the primary tumor (> 45mm), resection margin involvement, and perineural invasion were independent prognostic factors for resected PC in this cohort. This is the first study to report perineural invasion as an independent prognostic factor. In addition, our cohort contained patients who did not undergo resection but whose tumor gene expression status was known because intraoperative biopsies were performed. This allowed comparison between resected and nonresected patients stratified by gene status. Operative resection was associated with a longer survival in those patients with loss of DPC4/Smad4 expression in their tumor. In contrast, the survival of those patients whose tumors expressed DPC4/Smad4 was no different with resection compared with operative biopsy alone. DPC4/Smad4 expression may be a potentially important indicator of response to resection of patients with PC, as DPC4/Smad4 expression can be assessed without pancreatic resection. Aberrant expression of p53, p21WAF1/CIP1, p16INK4A, or cyclin D1 was not associated with differences in outcome in our cohort. Indicators of prognosis in PC are inconsistent and hence remain controversial. There are few large series of PC assessing the influence of clinicopathologic parameters and treatment on prognosis. Not all studies report similar findings. A summary of the most recently published significant series is presented in Table 8. PC, despite its high mortality rate, sometimes behaves in contrast to expectations based on clinicopathologic findings, with long-term survivors sometimes displaying poor prognostic features.11 There are also isolated case reports of patients with PC surviving up to 12 years without operative resection.34 Thus, the inaccuracy of these traditional clinicopathologic prognostic indicators and the difficulties in assessing them before resection may result in inappropriate patient selection for operative intervention. Molecular markers may become increasingly important not only as prognostic indicators but also for patient selection for treatment in PC, and they may serve as surrogate markers of the extent of disease and response to therapy. Further investigation of DPC4/Smad4 expression as one such marker in larger cohorts receiving optimal therapy and preoperative assessment is encouraged.
The actuarial 5-year survival rate for our cohort for all resected tumors was 11% compared with other series, which have reported 5-year survival rates of 6.8%,5 17%,3 and 24%.4 Most series, including our series, report primary tumor size as an independent prognostic factor but differ with respect to the effect of nodal status, margin involvement, perineural invasion, and adjuvant therapy (Table 5).3-6,35 The variability in the prognostic influences of these parameters may be attributed in part to differences in the proportions of patients exhibiting these characteristics at different centers (Table 8), which may reflect patient selection for resection. There is also a degree of subjectivity in assessing parameters such as degree of differentiation, microscopic margin status, and perineural invasion. The high mortality and rapid demise of patients with PC may also influence statistical analysis. In our cohort of resected tumors, the proportions and influence of tumor size and lymph node status were comparable to those reported in the literature (Table 5). However, perineural invasion by tumor was present in 80% of specimens compared with 21%4 and 12%,5 despite strict guidelines for determining perineural invasion in our study. The reason for this is unclear. DPC4/Smad4 expression within the PC of resected pancreata in our cohort was associated with a worse outcome on univariate analysis. Multivariate analysis (Table 6) revealed that DPC4/Smad4 expression was related to tumor size and to lymph node involvement (although lymph node involvement was more strongly related to tumor size than DPC4/Smad4 expression). DPC4/Smad4 expression was independent of margin status and perineural invasion in patients who underwent resection. The advantage of DPC4/Smad4 expression as a prognostic indicator is that it is potentially assessable preoperatively or during staging laparoscopy, whereas other factors, such as tumor size, resection margins, perineural invasion, and lymph node status, are determined after resection. Its lack of independence as a prognostic marker because of its relationship to tumor size and lymph node status suggests that DPC4/Smad4 expression may be a surrogate marker of these prognostic factors, which are difficult to determine without resection. Immunolabeling for DPC4/Smad4 expression is reproducible and reliable and correlates strongly with DPC4 gene status determined by alternative means.30 A recent series, and the only previously published series of DPC4/Smad4 expression and prognosis in PC, reported that loss of DPC4/Smad4 expression was a minor, yet independent, poor prognostic factor (hazard ratio = 1.36, 95% confidence interval = 1.01 to 1.83; P = .04) in 250 resected pancreata.33 Loss of DPC4/Smad4 expression within the invasive carcinoma of all patients in our cohort was associated with tumor resectability and hence conferred a better prognosis overall on univariate but not multivariate analysis. Multivariate analysis of all patients in our cohort (n = 129) also demonstrated that tumor stage and degree of differentiation were independent of operative treatment (resection/operative biopsy). Although "resectability" is multifactorial, this analysis suggests that better methods of selecting patients for resection may be developed. The association of DPC4/Smad4 expression with a poor outcome is perhaps contrary to what might be expected to occur with loss of expression of a tumor suppressor gene. The relationship of loss of DPC4/Smad4 expression to tumor size, lymph node metastases, and resectability suggests that those tumors with detectable loss of DPC4/Smad4 expression somehow resulted in earlier presentation and hence more effective treatment. In addition, there are now a number of examples of known tumor suppressor genes whose loss or gain of expression cosegregates with a worse outcome.36,37 Experimental models have shown that transient overexpression of DPC4/Smad4 induces p21WAF1/CIP1 transcription and leads to growth inhibition in some cell systems.21 Although p21WAF1/CIP1, the downstream target of Smad4, plays a role in inhibiting cyclin E/cdk2 activity that results in cell-cycle arrest, it also promotes the assembly of cyclin D1/cdk4 complexes that promote cell-cycle progression.38 Interestingly, overexpression of p21WAF1/CIP1 is an early event in the development of pancreatic intraepithelial neoplasia, believed to be ductal precursor lesions of PC.39 Only a moderate reduction in proliferation and induction of apoptosis is due to Smad4 nuclear translocation,40 and Smad4-independent TGF-ß–mediated cell-cycle inhibition through ERK inactivation is a more potent mechanism for cell-cycle inhibition and induction of apoptosis.41 There is mounting evidence that the role of the TGF-ß signaling pathway in cancer is contextual and that although reintroduction of DPC4/Smad4 in cell lines leads to growth inhibition, stably transfected DPC4/Smad4 colon cancer cell lines demonstrated increased motility.42 Furthermore, loss of TGF-ß signaling in right sided colon cancers was associated with a decrease in the rate of lymph node metastasis.43 As DPC4/Smad4 seems to exert its tumor suppressor effect by inducing transcription of the cdk inhibitor p21WAF1/CIP1, one would expect that if p21WAF1/CIP1 expression was maintained in the absence of DPC4/Smad4 expression, ie, p21WAF1/CIP1 expression was not dependent on DPC4/Smad4, then its functional role in suppression of proliferation may be abrogated while its other potential effects (increased motility and metastatic potential) could remain. Our data reveal that those tumors that exhibited loss of DPC4/Smad4 expression but had high levels of p21WAF1/CIP1 expression conferred a higher resectability rate than those with loss of DPC4/Smad4 expression alone. This group also demonstrated a survival advantage on univariate analysis, but this was not better than loss of DPC4/Smad4 expression alone. Hence, it is possible that those cancers that develop through aberrations in DPC4 manifest as a less aggressive phenotype, especially if the levels of p21WAF1/CIP1 are maintained by other mechanisms. Heterozygosity for DPC4/Smad4 may be more important than homozygous loss as DPC4/Smad4 heterozygous mice develop gastric, duodenal, and colonic polyps that progress to invasive cancer. Only a small proportion of these exhibit loss of DPC4/Smad4 protein expression44,45; however, the effect on the pancreas is unknown. In addition, loss of heterozygosity studies suggest that heterozygosity within the region of the DPC4 locus exists in early pancreatic intraepithelial neoplasia lesions,46 whereas loss of DPC4/Smad4 expression occurs later in the development of PC. Immunohistochemical analysis may not differentiate between wild-type DPC4/Smad4 expression and mutant DPC4/Smad4 proteins that may have different functional consequences. These findings suggest that functions attributed to these gene products in cell culture models cannot always predict their observed behavior in vivo. Our study also determined that aberrant expression of p53, p21WAF1/CIP1, p16INK4A, or cyclin D1 is not associated with prognosis, which is in agreement with a recently reported cohort of PC patients.6 Despite some reports suggesting that differential expression of cyclin D114 and p16INK4A 47 cosegregate with prognostic groups in small studies, the evidence is mounting against these possibilities. In conclusion, prognostic factors for PC, with the exception of tumor size, remain inconsistent and poorly defined, are difficult to accurately determine preoperatively, and on the whole depend on examination of the resected pancreas. We have identified perineural invasion as an independent prognostic factor together with tumor size and resection margin involvement in resected pancreata. Assessment of DPC4/Smad4 expression may provide an alternative method or surrogate marker for estimating the prognosis of patients with PC because it can be determined without resection. Further study of the influence of DPC4/Smad4 expression in pancreatic adenocarcinoma in large cohorts who undergo optimal therapy is essential to its further evaluation as a prognostic and therapeutic marker in PC.
Supported by grants from the National Health and Medical Research Council of Australia (NHMRC), Royal Australasian College of Surgeons, R.T. Hall Trust, The Cancer Council New South Wales, St Vincent’s Hospital, and St Vincent’s Clinic Foundation. A.V.B. is the holder of an NHMRC Medical Postgraduate Research Scholarship and the recipient of the Sir Roy McCaughey Fellowship from the Royal Australasian College of Surgeons. The authors acknowledge the contributions of Alexander Condoleon and Catherine Langusch in data management and the staff of the pathology and clerical services at St Vincent’s Hospital Campus, Westmead Hospital, and Royal Prince Alfred Hospital for their assistance in data acquisition.
1. Hruban RH, Yeo CJ, Kern SE: Pancreatic cancer, in Vogelstein B, Kinzler KW (eds): The Genetic Basis of Human Cancer. New York, NY, McGraw-Hill, 1996, pp 603-613 2. Kimura W, Morikane K, Esaki Y, et al: Histologic and biologic patterns of microscopic pancreatic ductal adenocarcinomas detected incidentally at autopsy. Cancer 82: 1839-1849, 1998[CrossRef][Medline] 3. Sohn TA, Yeo CJ, Cameron JL, et al: Resected adenocarcinoma of the pancreas—616 patients: Results, outcomes, and prognostic indicators. J Gastrointest Surg 4: 567-579, 2000[CrossRef][Medline] 4. Geer RJ, Brennan MF: Prognostic indicators for survival after resection of pancreatic adenocarcinoma. Am J Surg 165: 68-73, 1993[CrossRef][Medline] 5. Nitecki SS, Sarr MG, Colby TV, et al: Long-term survival after resection for ductal adenocarcinoma of the pancreas: Is it really improving? Ann Surg 221: 59-66, 1995[Medline] 6. Kawesha A, Ghaneh P, Andren-Sandberg A, et al: K-ras oncogene subtype mutations are associated with survival but not expression of p53, p16(INK4A), p21(WAF-1), cyclin D1, erbB-2 and erbB-3 in resected pancreatic ductal adenocarcinoma. Int J Cancer 89: 469-474, 2000[CrossRef][Medline]
7. Sherr CJ: Cancer cell cycles. Science 274: 1672-1677, 1996
8. Rozenblum E, Schutte M, Goggins M, et al: Tumor-suppressive pathways in pancreatic carcinoma. Cancer Res 57: 1731-1734, 1997
9. Kirsch DG, Kastan MB: Tumour-suppressor p53: Implications for tumour development and prognosis. J Clin Oncol 16: 3158-3168, 1998 10. Barton CM, Staddon SL, Hughes CM, et al: Abnormalities of the p53 tumour suppressor gene in human pancreatic cancer. Br J Cancer 64: 1076-1082, 1991[Medline] 11. DiGiuseppe JA, Yeo CJ, Hruban RH: Molecular biology and the diagnosis and treatment of adenocarcinoma of the pancreas. Adv Anatom Pathol 3: 139-155, 1996
12. Velculescu VE, El-Deiry WS: Biological and clinical importance of the p53 tumour suppressor gene. Clin Chem 42: 858-868, 1996 13. Hunter T, Pines J: Cyclins and cancer II: Cyclin D and CDK inhibitors come of age. Cell 79: 523-582, 1994
14. Gansauge S, Gansauge F, Ramadani M, et al: Overexpression of cyclin D1 in human pancreatic carcinoma is associated with poor prognosis. Cancer Res 57: 1634-1637, 1997 15. Liggett WH, Sidransky D: Role of the p16 tumour suppressor gene in cancer. J Clin Oncol 16: 1197-1203, 1998[Abstract] 16. Ruas M, Peters G: The p16INK4A/CDKN2A tumor suppressor and its relatives. Biochim Biophys Acta 1378: F115-177, 1998[Medline] 17. Caldas C, Hahn SA, da Costa LT, et al: Frequent somatic mutations and homozygous deletions of the p16 (MTS 1) gene in pancreatic cancer. Nat Genet 8: 27-32, 1994[CrossRef][Medline]
18. Schutte M, Hruban RH, Geradts J, et al: Abrogation of the Rb/p16 tumor-suppressive pathway in virtually all pancreatic carcinomas. Cancer Res 57: 3126-3130, 1997
19. Piek E, Heldin CH, Ten Dijke P: Specificity, diversity and regulation in TGF-beta superfamily signaling. FASEB J 13: 2105-2124, 1999
20. Grau AM, Zhang L, Wang W, et al: Induction of p21waf1 expression and growth inhibition by transforming growth factor beta involve the tumor suppressor gene DPC4 in human pancreatic adenocarcinoma cells. Cancer Res 57: 3929-3934, 1997
21. Hunt KK, Fleming JB, Abramian A, et al: Overexpression of the tumour suppressor gene Smad4/DPC4 induces p21 expression and growth inhibition in human carcinoma cells. Cancer Res 58: 5656-5661, 1998 22. Villanueva A, Garcia C, Paules AB, et al: Disruption of the antiproliferative TGF-beta signaling pathways in human pancreatic cancer cells. Oncogene 17: 1969-1978, 1998[CrossRef][Medline] 23. Sweeney KJ, Swarbrick A, Sutherland RL, et al: Lack of relationship between CDK activity and G1 cyclin expression in breast cancer cells. Oncogene 16: 2865-2878, 1998[CrossRef][Medline] 24. Sutherland RL, Watts CKW, Lee CSL, et al: Breast cancer, in Masters JRW, Palsson B (eds): Human Cell Culture. London, United Kingdom, Kluwer Academic Publishers, 1999, pp 79-106 25. Carroll AG, Voeller HJ, Sugars L, et al: p53 oncogene mutations in three human prostate cancer cell lines. Prostate 23: 123-134, 1993[Medline]
26. Pardali K, Kurisaki A, Moren A, et al: Role of Smad proteins and transcription factor Sp1 in p21Waf1/Cip1 regulation by transforming growth factor-beta. J Biol Chem 275: 29244-29256, 2000
27. Migaldi M, Sgambato A, Garagnani L, et al: Loss of p21Waf1 expression is a strong predictor of reduced survival in primary superficial bladder cancers. Clin Cancer Res 6: 3131-3138, 2000
28. Bova RJ, Quinn DI, Nankervis JS, et al: Cyclin D1 and p16INK4A expression predict reduced survival in carcinoma of the anterior tongue. Clin Cancer Res 5: 2810-2819, 1999 29. Alle KM, Henshall SM, Field AS, et al: Cyclin D1 protein is overexpressed in hyperplasia and intraductal carcinoma of the breast. Clin Cancer Res 4: 847-854, 1998[Abstract]
30. Wilentz RE, Su GH, Le Dai J, et al: Immunohistochemical labeling for Dpc4 mirrors genetic status in pancreatic adenocarcinomas: A new marker of DPC4 inactivation. Am J Pathol 156: 37-43, 2000
31. Wilentz RE, Geradts J, Maynard R, et al: Inactivation of the p16 (INK4A) tumor-suppressor gene in pancreatic duct lesions: Loss of intranuclear expression. Cancer Res 58: 4740-4744, 1998 32. Sobin LH, Wittekind C: TNM Classification of Malignant Tumours: UICC International Union Against Cancer, ed 5 . New York, NY, Wiley-Liss, 1997
33. Tascilar M, Skinner HG, Rosty C, et al: The SMAD4 protein and prognosis of pancreatic ductal adenocarcinoma. Clin Cancer Res 7: 4115-4121, 2001 34. Silberstein E, Walfish S, Lupu L, et al: Twelve-year survival after the diagnosis of locally advanced carcinoma of the pancreas: A case report. J Surg Oncol 75: 142-145, 2000[CrossRef][Medline] 35. Fortner JG, Klimstra DS, Senie RT, et al: Tumor size is the primary prognosticator for pancreatic cancer after regional pancreatectomy. Ann Surg 223: 147-153, 1996[CrossRef][Medline]
36. Henshall SM, Quinn DI, Lee CS, et al: Overexpression of the cell cycle inhibitor p16INK4A in high grade prostatic intraepithelial neoplasia predicts early relapse in prostate cancer patients. Clin Cancer Res 7: 544-550, 2001 37. Dong Y, Walsh MD, McGuckin MA, et al: Increased expression of cyclin dependent kinase inhibitor 2 (CCDKN2A) gene product p16INK4A in ovarian cancer is associated with progression and unfavourable prognosis. Int J Cancer 74: 57-63, 1997[CrossRef][Medline]
38. LaBaer J, Garrett MD, Stavenson LF, et al: New functional activities for the p21 family of CDK inhibitors. Genes Dev 11: 847-862, 1997
39. Biankin AV, Kench JG, Morey AL, et al: Overexpression of p21WAF1/CIP1 is an early event in the development of pancreatic intraepithelial neoplasia. Cancer Res 61: 8830-8837, 2001
40. Le Dai J, Bansal RK, Kern SE: G1 cell cycle arrest and apoptosis induction by nuclear Smad4/Dpc4: Phenotypes reversed by tumorigenic mutation. Proc Natl Acad Sci U S A 96: 1427-1432, 1999 41. Giehl K, Seidel B, Gierschik P, et al: TGF-beta1 represses proliferation of pancreatic carcinoma cells which correlates with Smad-4 independent inhibition of ERK. Oncogene 19: 4531-4541, 2000[CrossRef][Medline] 42. Schwarte-Waldhoff I, Klein S, Blass-Kampmann S, et al: DPC4/SMAD4 mediated tumor suppression of colon carcinoma cells is associated with reduced urokinase expression. Oncogene 18: 3152-3158, 1999[CrossRef][Medline] 43. Iacopetta BJ, Welch J, Soong R, et al: Mutation of the transforming growth factor-beta type II receptor gene in right sided colorectal cancer: Relationship to clinicopathological features and genetic alterations. J Pathol 184: 390-395, 1998[CrossRef][Medline]
44. Takaku K, Miyoshi H, Matsunaga A, et al: Gastric and duodenal polyps in Smad4 (Dpc4) knockout mice. Cancer Res 59: 6113-6117, 1999 45. Xu X, Brodie S, Yang X, et al: Haploid loss of the tumor suppressor Smad4/Dpc4 initiates gastric polyposis and cancer in mice. Oncogene 19: 1868-1874, 2000[CrossRef][Medline]
46. Heinmoller E, Dietmaier W, Zirngibl H, et al: Molecular analysis of microdissected tumors and preneoplastic intraductal lesions in pancreatic carcinoma. Am J Pathol 157: 83-92, 2000 47. Naka T, Kobayashi M, Ashida K, et al: Aberrant p16(INK4A) expression is related to clinical stage and prognosis in patients with pancreatic cancer. Int J Oncol 12: 1111-1116, 1998[Medline] Submitted December 14, 2001; accepted August 8, 2002. This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
INTRODUCTION
