2005 The Clinical Significance of p53 Accumulation in Gastric Carcinoma Teresa Starzynska, M.D., Ph.D.l Maciej Markiewski, M.D., Ph.D.2 Wenancjusz Domagala, M.D., Ph.D.2 Krzysztof Marlicr, M.D., Ph.D.’ Janusz Mietkiewski, M.D? Stephen A. Roberts, Ph.D.4 Peter L. Stern, Ph.D.S ’ Department of Gastroenterology, Medical Pomeranian Academy, Szczecin, Poland. Department of Pathology, Medical Pomeranian Academy, Szczecin, Poland. Department of Pathology, General Hospital, Szczecin-Police, Poland. Cancer Research Campaign Biomathematics and Computing Unit, f’aterson Institute for Cancer Research, Christie Hospital (NHS) Trust, Manchester, United Kingdom. Cancer Research Carnpaign Department of Immunology, Paterson Institute for Cancer Research, Christie Hospital (NHS) Trust, Manchester, United Kingdom. Presented in part at ihe 2nd United European Gastroenterology Week, Barcelona, Spain, July 2, 1993; at the 3rd United Gastroenterology Week, Oslo, Norway, June 1994; and at the 10th World Congress of Gastroenterology, Los Angeles, CA, 1994. Supported by the Cancer Research Campaign of Great Britain, Polish Foundation of Gastroenterology, Pomeranian Medical Academy, Szczecin, Poland, and BritisWPolish Joint Research Collaboration Programme. The authors thank Professor David P. Lane for the gift of CMI antiserim used in these studies. Address for reprints: Peter L. Stern, Ph.D., Department of Immunology, Paterson Institute for Cancer Research, Marichester M20 9BX, U.K. Received October 27, 1995; revision received February 5, 1996; accepted February 5, 1996. 0 1996 American Cancer Society BACKGROUND. Alterations in the expression of p53 tumor suppressor protein is a frequent event in human cancer but the practical implications of this phenomenon are yet to be fully exploited. The objective of this study was to determine the value of p53 accumulation as a marker of tumor progression and prognosis of gastric carcinoma patients and to evaluate whether this parameter can be properly assessed prior to surgery. METHODS. The expression of p53 was studied immunohistochemically in 200 gastric carcinomas using paraffin embedded surgical specimens and endoscopic biopsies. The correlation between p53 expression in tumor tissue, selected clinicopathologic variables, and the course of the patients’ disease were analyzed. RESULTS. Results showed that 42.5% of the gastric carcinomas expressed elevated levels of p53 protein. P53 accumulation positivity correlated with increasing tumor stage and size ( P < 0.001 and P = 0.025, respectively). P53 positive tumors had a higher propensity for lymph node and distant metastases ( P < 0.001). P53 accumulation was also more frequently detected in carcinoma from proximal rather than distal stomach ( P = 0.027). In patients receiving potentially curative resection for advanced cancer, p53 accumulation was an independent parameter and the strongest for poor prognosis (RR = 3.7, P < 0.001). There was complete concordance between immunohistochemical detection of p53 in endoscopic and surgical material. CONCLUSIONS. A preoperative assessment of p53 expression in gastric carcinoma can be helpful to identify patients at high risk of metastatic spread to regional lymph nodes and independently to identify those with especially poor prognosis. When combined with routine procedures, this simple and inexpensive test might allow appropriate planning of better treatment strategies. Cancer 1996; 77:2005-12. 0 1996 American Cancer Society. KEYWORDS: gastric carcinoma, p53 protein, immunohistochemistry, prognosis. M utations of the tumor-suppressor p53 gene on chromosome 17p are a very frequent event in many human cancers but the clinical usefulness of this phenomenon remains to be defined.’-4 In normal cells and tissue, p53 protein has a very short half-life and attains such a low In contrast, the level that it is not detectable immunohistochemically.s~6 altered forms, present in 30 to 80% of different neoplasms, are more stable and accumulate to concentrations that can be detected by immunohisto~hernistry.~-” Thus a simple immunohistochemical procedure which stratifies malignant tumors into those with and without p53 accumulation can be analyzed for useful correlations with clinical and histopathologic data. Gastric carcinoma remains an important medical problem. Although the incidence of the disease is generally declining, it is still the second Despite progress in diagleading cause of cancer death in the 2006 CANCER May 15,1996 / Volume 77 I Number 10 nostic techniques, in countries other than Japan, the overall 5-year survival rate is less than 20%."-'4 Surgery is the treatment of choice for this cancer, but the majority of patients will suffer a relapse following operation. Therefore, different treatment modalities, such as gastric resection with extensive lymphadenectomy and an aggressive chemotherapy, have been investigated.",'"''" Nevertheless, not all patients benefit from this kind of treatment. The selection of the most appropriate surgical and chemotherapeutic approach is clearly central for maximizing patient disease management. This selection is currently based primarily on TNM staging. However, for gastric carcinoma, tumor staging, particularly regional lymph node status, is very difficult to establish before surgery and the subsequent histologic examination of resected specimens. Furthermore, TNM classification does not give a complete separation of good and bad prognosis for individual patients. New markers which would improve the accuracy of preoperative assessment of gastric carcinoma stage and its aggressiveness are clearly required. The clinical significance of p53 accumulation in gastric carcinoma remains a controversial issue. While some authors have found this alteration to be associated with advanced stages of the disease and shorter survival time, others detected no such orr relation."^^' In a previous study of 55 gastric carcinoma patients, we demonstrated that the presence of p53 accumulation in tumor tissue might favor a more aggressive clinical course of the disease." The objective of the present study was to determine, on a large series of patients, the value of p53 overexpression as a marker of tumor progression and prognosis in this cancer and to evaluate whether this parameter can be properly assessed prior to surgery. METHODS Gastric Carcinoma Patients Two hundred patients with gastric carcinoma, diagnosed and treated between 1987 and 1994 at the Departments of Gastroenterology, Pathology, and the First Department of Surgery, Pomeranian Medical Academy, Szczecin, Poland were included in this study. All cases had adequate clinical and pathologic information. Patient's sex, age, family history of cancer, durations of symptoms, tumor site, size, stage, and histology, type of treatment performed, and follow-up data were recorded. The median age of the patients was 59 years (range: 24-90 years) and 67.5% of the patients in the series were male. Sixty one gastric tumors (30.5%)were located in the antrum or pylorus region, 85 (42.5%) in the body, 43 (21.5%) in the cardia or fundus, in 4 patients (2%) carcinoma affected nearly the whole stomach, and in 7 patients (3.5%)cancer developed in the gastric stump. For further analysis, gastric carcinomas were divided into proximal and distal le- sions. Stage and histologic type of gastric carcinoma were assessed from routine examination of paraffin sections obtained from surgical specimens, stained with hematoxylin and eosin (H & E). Evaluation of tumor stage was performed according to the criteria of the International Union Against Cancer (UICC) additionally tumor were classified as early or advanced according to the criteria of the Japanese Research Society for Gastric Histologic typing was assessed according to Lauren's clas~ification.'~A potentially curative resection was performed on 128 patients, 26 underwent noncurative gastric resection, and 46 underwent palliative surgical procedures or laparatomy. Patients were asked to continue their clinical care at the Department of Gastroenterology in Szczecin. Control visits were made every 3 months for the first 2 years after surgery and then yearly. These visits included conventional clinical examination, ultrasonography of the abdomen, and gastroscopy. Patients could come to the clinic independently of the schedule if they felt unwell. Time of recurrence or metastases, survival time, and cause of death were recorded. Tissue Preparation All of the patients were examined endoscopically before operation, and a minimum of four tissue specimens were sampled with routine biopsy forceps. These specimens were fixed in 80% ethanol and embedded in paraffin. Surgical material from the same patients was fixed in 10% neutral formalin and also embedded in paraffin. Surgical material from all 200 patients was used for immunohistochemical analyses. Additionally, in 90 patients, p53 expression was studied using endoscopic material. lmmunohistochemistry Histology of all of the cases was reviewed and representative sections from 2 to 5 different areas of each tumor were selected for immunohistochemistry. These were deparaffinized in xylene, transferred to 95% ethanol, and air-dried. The polyclonal rabbit anti-p53 antibody, CM1, raised against the full length human p53 protein, was used as a primary antibody. A 3-stage immunoperoxidase technique was used as previously described." Briefly, the slides were incubated overnight at 4 "C with polyclonal anti-p53 antibody and diluted 1/750. The sections were washed and treated consecutively for 30 minutes with biotinylated swine anti-rabbit antibody (Dakopatts, Glostrup, Denmark) diluted 1/400 in TBS and streptavidin HRP-conjugated reagent (Dako Ltd., U.K.) diluted 1/BOO, at room temperature. Peroxidase was visualized using 0.05% solution of diaminobenzidine tetrahydrochloride (Sigma, St. Louis, MO) and 0.1% solution of nickel chloride in TBS containing 0.03% hydrogen peroxide (10 min). Finally, the slides were lightly stained in haemalum (Mayer's, Merck, Poole, UK). Replacement of the primary anti- p53 Protein in Gastric Carcinorna/Starzynska et al. 2007 body with normal rabbit serum was used as a negative control. Colonic carcinoma sections with p53 accumulation were includled in each experiment to ensure that the procedure was working optimally. Specific nuclear staining of cells (regardless of proportion) was considered positive for abnormal p53 expression. Statistical Analysis For all 200 patients, p53 expression in gastric carcinoma was compared with selected clinicopathological variables. These included sex, age, family history of cancer, duration of symptoms, tumor site, size, depth, histologic type, and metastatic status at presentation. Comparisons between p53 expression in tumor tissue and these parameters was performed using the chi-square test and Fisher's exact test. Among 128 patients who undeiwent potentially curative ga:stric resection, 111 were included in the survival analysis. The median follow up time for these patients was 33 months (range: 3-85 mos). The disease free survival time was defined as the time to cancer recurrence. The overall survival time was defined as the time to cancer related death. Univariate survival analysis was performed using; the log rank test for a whole group of patients after potentially curative gastric resection (n = 111; including 19 patients with early tumors) and independently for a subgroup of patients with advanced cancer (n = 92).2R.L'? The variables assessed were: age (550 vs. >50 years), sex, duration of symptoms ( < G vs. >6 mos), tumor site (distal vs. proximal stomach), size ( < 5 vs. >5 cm), depth, histology (intestinal vs. diffuse), regional lymph node status, and p53 expression in cancer tissue. Multivariate Cox regression was used to determine which parameters were independently of prognostic iniportance.'" There were no cancer-related deaths in the 19 patients with early disease; the outcome of these patients is completely determined by the 1 parameter, therefore these patients were excluded from multivariate analysis. A step-up procedure was used to build the model, including only terms that were significant at the 0.05 level. Significance levels quoted are those to remove each of the terms in the final model from that model. A significance level of 0.05 was used throughout. RESULTS The Expression of p53 in Gastric Tissue p53 accumulation was detected in 85 out of 200 (42.5%) gastric carcinomas. Irnmunoreaction was always localized in the nuclieus of neoplastic cells (Fig. 1).A negative reactivity with anti-p53 antibody was seen in gastric epithelium adjacent to carcinoma, including intestinal metaplasia (11 = 154) and in stroma (n = 200). There was complete concordance between the final results of immunohistochemical staining among 90 carcinomas stained, FIGURE 1. lmmunohistochemical detection of p53 accumulation 111 en- doscopic biopsies is shown. Gastric carcinoma of a well differentiated intestinal type(a) and a poorly differentiated tubular carcinoma (b) shs3wing strong nuclear staining (original magnification (a) x180; (b) ~ 4 0 0 ) . using both endoscopic and surgical material. Thirty-eight of these tumors showed positive and 52 negative reactivity. However, the intensity and pattern of immuno renction were related to the type of material used. On endoscopic biopsies, p53 staining was always much stronger and distributed in a greater proportion of the cells. Forty seven percent and 32% of p53 positive endoscopic biopsies exhibited labelling in more than 95% and betweim 50 and 75% of tumor cells, respectively. In the remaining 21%, p53 accumulation was detected in less than 50% of the cancer cells, however, no case with less than 10% of positive cells was seen. By contrast, 29 out of 38 (76%) of the surgical specimens, scored as p53 positive, exhibited focal reactivity. In no case were only occasional positive cells found, although some specimens showed nonreactive areas especially in the centre of the section. The proportions of p53 positive cells and the intensity of staining could vary between different areas of the same tumor but the pattern was independent of tumor depth. The 2008 CANCER May 15,1996 / Volume 77 / Number 10 preservation of p53 antigenicity in endoscopic specimens was probably improved by the rapid fixation offered by 80% alcohol and small biopsy size. TABLE p53 Overexpression and Selected Clinicopathologic Features The relationships between p53 expression in gastric cancer and selected clinicopathologic variables are summarized in Table 1. There was a highly significant correlation between the presence of p53 accumulation in neoplastic tissue and tumor stage. The proportion of p53 positively reacting carcinomas increased as the stage progressed ( P < 0.001). Only 15% of the gastric carcinomas from the patients without metastases were p53 positive compared with 50% of the tumors which had invaded regional lymph nodes and 67% of the neoplasms with distant metastases ( P < 0.001). Twenty-two percent, 28%, 37%, and 62% of T1, T2, T3, and T4 carcinomas, respectively, overexpressed p53 ( P < 0.001). Fewer of the tumors classified as early showed p53 overexpression compared with tumors classified as advanced (19%vs. 45%; P = 0.039). p53 overexpression was detected more frequently in carcinoma from proximal rather than distal stomach (60%and 38%, p53 overexpressing tumors, respectively; P = 0.027). There was also an association between p53 positivity and tumor size, with 25% of the small (<2 cm) tumors being p53 positive compared with 52% (2-5 cm) and 41% (>5 cm) of the larger tumors ( P = 0.025). p53 expression was unrelated to patient sex, age, family history of cancer, duration of symptoms, and histologic type. Parameter Relationship between p53 Expression and Survival Of 111 patients who underwent potentially curative surgery, including 19 with early and 92 with advanced gastric carcinoma, 39 had tumors with p53 accumulation and 72 had carcinoma with undetectable p53 protein. Figure 2 shows the actuarial survival curves of this group, and of the p53 negative and positive subgroups. In these patients, the presence of p53 accumulation was a marker of poor prognosis. The overall 5-year survival rates were 20% for patients with carcinoma overexpressing the p53 protein and 70% for patients who had tumors with an undetectable level of p53 ( P < 0.001). The analogous results for disease free survival time were 28% versus 72% ( P < 0.001). Results of univariate analyses (Table 2) show that in addition to p53 accumulation, the prognosis of these patients was significantly adversely affected by metastatic spread to the regional lymph nodes ( P < 0.001) and increasing tumor depth ( P < 0.001).There was a weak effect on the disease free interval with increasing duration of symptoms ( P = 0.0171, but this was not carried through to survival. The other indicators were not significantly associated with survival or disease free interval. 1 Relationship between p53 Expression in Gastric Carcinoma and Selected Clinicopathologic Features Number examined p53 accumulation positive Sex male female 135 65 59 (44%) 26 (40%) NSa Age <50 >50 53 147 18 (34%) 67 (46%) NS Family history of cancer Positive Gastric cancer Other cancers Negative No reply 61 32 29 98 41 23 (38%) 12 (38%) 11 (38%) 39 (40%) NS 51 54 63 30 25 (49%) 24 (44%) 27 (43%) 7 (23%) NS 43 146 11 26 (60%) 55 (38%) 4 (36%) 36 77 87 9 (25%) 40 (52%) 36 (41%) P = 0.025 114 86 49 (43%) 36 (42%) NS 22 61 41 68 5 (22%) 17 (28%) 15 (37%) 42 (62%) P < 0,001 Early cancer Advanced cancer 21 179 4 (19%) 81 (45%) P = 0.039 Metastases at presentatione Negative Only lymph node Distant 68 80 51 10 (15%) 40 (50%) 34 (67%) Duration of symptoms (mosIb 0-6 7-12 13-60 3 0 Tumor site Proximal stomach Distal stomach Other' P = 0.027 Tumor size (cm) <2 2-5 >5 Histology Intestinal Diffuse Tumor depthd T1 T2 T3 T4 P < 0.001 NS: not significant. a Not significant, ,yz test. Data available from 198 patients. 'Gastric stump or nearly all stomach affected. From 192 patients. ' From 199 patients. p53 Protein in Gastric CarcinomalStarzynska et al. lymph node status achieved statistical significance ( P < 0.001; RR -= 3.8 and P = 0.035, HI< -= 2.4, respectively). b a 100 "j ..... 90 ..... ....... 70 .\" .-2 I CONCLUSIONS .. . ... .. . . 80 ....,,..... .......14172 . 60 40 5 b , ; ; ' ; , , , , v) 30 28139 20 t 10 1 0 0 20 40 2009 60 Time (Months) 80 100 0 . I 20 . I 40 . I . 60 I 80 . I 100 Time (Months) FIGURE 2. Acturial survival curves for p53 negative (broken line) and positive (solid line) platients after potentially curative resection for gastric carcinoma are shown: (a) all patients. (b) patients with advanced disease. The numbers 01 dearhs and patients in each group are given. Similar results were observed for the 92 patients who underwent potentially curative surgery for advanced gastric carcinoma. Figure 2b shows survival curves of this group, and for I he p53 positive and negative subgroups. I n these patienrs, the presence of p53 accumulation was a marker of poor prognosis. The overall 5-year survival rates were 15% for patients with carcinomas overexpressing 1153 protein and 63% for patients who had tumors with undetectable levels of p53 (1' < 0.001). The analogous results of disease free survival tinie were 23% versus 65% (I' <: 0.001). The results of univariate analysis show that the prognosis of these patients was also significantly adversely affected by metastatic spread to regional lymph nodes ( P < 0.001) and increasing tumor depth ( P < 0.0L2). The other indicators were not significantly associated with survival or disease free interval. 'l'ablc 2 also shows that there were no cancer-related deaths in the I!>patients with early disease; the outcome of these patients is completely determined by the 1 parameter thcrefore they were excluded from the multivariate analysis. I n the multivariate Cox regression analysis (?'able 3) of overall survival time for the group of patients after curative sui-gery for advanced carcinoma, the accumulation of p53 was the strongest predictor of poor prognosis ( I > < 0.001; relative risk (RK) = 3.7). The other parameters which achieved statistical significance were metastases to the regiona'l lymph nodes ( P = 0.025; HR = 2.5) and tumor depth ( P - - 0.012; RR = 2.2, for tumors without vs. with serosal involvement). IJsing cancer recurrence as the end point, similar results were obtained but only p53 and There is growing interest i n the supplementation of the traditional methods used to assess cancer progres:;ion and predict its course by use of different molecular markers.'9 Recent studies have suggested that the accumulation of p53 protein in neoplastic cells might be such a marker."" '" However, in contrast t o numerous papers discussing the clinical significance of this phenomenon, it has not been determined whether tumor samples obtained during diagnostic procedures constitute sufficient material to properly assess p.53 status. The problem is especially important in terms of heterogeneity of p53 accumulation in malignant tumors observed by many authors.l: 2lI.22 We have compared 1153 irnmunostaining on surgical and endoscopic material and clearly demonstrated that the presence of p53 accumulation in gastric carcinoma can be detected using endoscopic biopsy prior to surgery. We have also shown that this simple test can be helpful to predict tumor progression. In this study, the accuniulation of p53 protein was seen to occur in early gastric cancers confined to the mucosa but the proporticin of tumors exhibiting p53 overexpression increased as the tumor stage progressed. Thus, 22%, 28%, 37%. and 62% of T I , T2, T3, and '1.4 tumors, respectively, expressed elevated levels of p53. l'he analysis of our data also revealed a significant association between p53 accumulation and metastatic spread. p53 was detected in 15% of gastric carcinomas without metastases, 50% of tumors which had invaded regional lymph nodes, and in 67% of neoplasms with distant metastases. A strong relationship between p53 accumulation and metastases, together with the possibility of its simple preoperative assessment, might be useful in clinical practice to enhance our ability to select patients with a high risk of lymph node involvement and plan optimal surgical strategies. p53 studies on gastric carcinoma have produced conflicting results as to the relationship between p53 expression and tumor stage. In two recent studies of 180 and 149 gastric carcinoma patients, no correlation between p53 positivity, metastatic spread, and depth of tumor penetration were seen."'."' By contrast and in agreement with our results, 4 other studies of 77, 96, 135, and 357 patients, reported a higher propensity for metastasis to the regional lymph nodes in gastric carcinoma with p53 l~;~lwl..l5 accumu~ation, I n this study, p53 accumulation was also correlated with other indicators of poor prognosis for gastric cancer being more frequently detected in carcinoma from proximal rather than distal stomach and in larger neoplasms. The Lauren classification divides gastric carcinoma 2010 CANCER May 15,1996 / Volume 77 I Number 10 TABLE 2 Univariate Log Rank Analyses of Prognostic Factors for 111 Gastric Carcinoma Patients (92 advanced tumors) Who Underwent Potentially Curative Surgery Feature p53 accumulation % 5-year 70 (63) 20 (15) 76 (69) 22 (17) sunival Sex M F Age 0-50 51 t 14/72 (141571 28139 (281351 10157 (10/42] 32/53 (32150) 0119 18152 (18152) 14125 (14125) 10115 (10115) 21152 (21139) 21/59 (21153) 8/20 (8117) 33/90 (33174) 26/71 (26152) 16/40 (16140) 30/74 (30161) 12/37 (12131) 13/33 (13127) 29178 (29165) 0-6 m >6 cm 10123 (10/21) 32188 (32/711 46 (44) 52 (42) None Stomach Other 17/53 (17144) 7/21 (7117) 7/19 (7115) 58 (50) 61 (55) 59 (48) Regional lymph node Metastases Tumor deplh Histology Tumor site Tumor size Duration of symptoms Family history of cancer -ve t ve -ve tve Ti T2 T3 T4 Deaths/ patients Diffuse Intestinal Proximal Distal 0-5 cm >5 cm 51 (54) 24 (24) 29 (29) 41 127) 56 (51) n!a 53 (44) 52 (38) 48 (48) 48 (38) 5 i (49) 46 (37) 52 (43) Parameter p53 accumulation Present vs. absent Lymph node metastases Positive vs. negative Tumor depth T2 vs. T3 t T4 Disease free Relative risk P value Relative risk P value 3.7 <0.001 3.8 <0.001 2.5 2.2 0.025 0.012 2.4 0.035 NS - NS: no1significant into intestinal and diffuse subtypes with different clinical, histologic, and epidemiologic characteristic^.^^ These differences suggest distinct etiologies and molecular mechanisms of development but in this study and those of others, p53 accumulation did not correlate with tumor histo~ogy,17,Z0,?I However, some authors observed p53 overexpression primarily in gastric carcinomas with tubular structure^.'^^^^ In spite of the vast array of established parameters disease free P value 0.96 (0.97) 72 (65) 28 (23) 75 (67) 32 (28) 100 51 (51) 43 (43) 43 (43) 56 (41) 56 (51) nla 60 (51) 55 (41) 59 (59) 51 (41) 66 (59) 59 (50) 55 (46) 0.12 (0.30) 39 (35) 60 (50) 0.017 (0.059) 0.81 (0.73) 68 (61) 52 (37) 57 (44) 0.56 (0.47) <0.001 1<0.001) <0.001 (<O.OOl) ID0 TABLE 3 Multivariate Analyses of Prognostic Factors for Patients after Potentially Curative Surgery for Advanced Cancer SUMVal % 5-year P value <O.OOI (0.012) 0.34 (0.075) 0.23 (0.37) 0.51 (0.61) 0.47 (0.41) <0.001 (<0.001) <0.001 (<0.001) 0.003 (0.18) 0.69 (0.25) 0.13 (0.22) 0.76 (0.43) 0.43 (0.37) 0.68 (0.69) which help to predict outcome of gastric carcinoma patients, individual tumor behavior still can not be accurately determined for the group of patients by TNM staging. Patients who are likely to survive (early carcinomas) or die (inoperable tumors) within 5 years can be identified. However, for patients with advanced, but operable gastric carcinoma, who constitute the largest group of patients with newly diagnosed disease, the likelihood of survival to 5 years is less certain. It is clearly important to identify those patients who are most likely to benefit from aggressive therapy. Our analysis showed a highly significant association between the presence of p53 accumulation and reduced survival time in a group of patients who underwent curative surgery for advanced gastric cancer. The overall and disease free 5-year survival rates for patients whose tumors accumulated p53 protein and in patients who had carcinoma with undetectable levels of p53 were 15%versus 63% and 23% versus 65%, respectively. Furthermore, p53 accumulation was an independent predictor and the strongest predictor of poor prognosis, followed by metastases to the regional lymph nodes, and serosal involvement. These results are consistent with a previous study of 76 patients which indicated that p53 accumulation can p53 Protein in Gastric CarcinomalStanynska et al. be an independent parameter of poor prognosis.” A significant positive correlation between p53 immunopositivity and shorter survival has been reported but the influence on prognosis for advanced but operable tumors was not determined.:’s Investigation of p53 status in early gastric carcinoma has shown increased incidence of p53 accumulation in patients with the poorest postoperative course3: and in those with tumors of a penetrating growth pattern.:’” The studies and work presented here are consistent with a role for p53 in expansion and aggressiveness of this malignancy. In conclusion, this report shows that in gastric carcinoma, the expression of p53 protein in tumor tissue can be properly assessed prior to surgery, using immunohistochemistiy on a small tissue samples obtained during endoscopy. Furthermore, we demonstrate that the preoperative assessmcnt of p53 expression in this carcinoma can be helpful in identifying patients at high risk for metastatic spread, including regional lymph node involvement, and in the discrimination of those patients with especially poor prognosis. When combined with routine procedures, this simple and inexpensive test might allow better planning of appropriate treatment strategies. REFERENCES I. 2. Lane DP.p53, guardian of the genome. Nurure 1992;358:156. Lane I>€’. Mutations of the p53 gene and accumulation of the p53 protein: common steps found in the majority of human cancers. In: fbriner JG, Rhoades JE, editors. Accomplishments in cancer research. Philadelphia: Jl3 Lippincott, 1990:252-66. Lane DP. The regulation of p53 function. Inf J C h c e r 1994;57523- 7. 4. Levine AJ, Momand J , Finlay CA. The p53 tumour suppressor gene. Nufirre 1991;351:453-6. 5. Finlay CA, Ilinds PW, Tan H-T, Elihyahu D, Oren M, Levine Aj. Activating mutations for transformation by p53 produce a gene product that forms an hsc 70-p53 complex with a n altered half life. Mol Cell Biol 1988;8:531-9. 6. Oren M. The p53 cellular tumor antigen: gene structure, expression and protein properties. Riochini Riophys Acfu 1985;823:67-78. 7. Gannon JV, Greaves R. lggo H, Lane DP. Activating mutations in p53 produce a common conformational effect. A monoclonal antibody specific for a mutant form. EMHO J 1990;9:1595- 1602. a. Iggo R, Gatter K, Bartek I, Lane DP, Harris AL. Increased expression of mutant forms of p53 oncogene in primary lung cancer. Lrtrzcet 1990;335:675-9. 9. Marks JR, Davidoff AM, Kerns HI, Humphrey PA, Pence JC, Dodge fM, et al. Overexpression and mutation of p53 in epithelial ovarian cancer. Cancer Res 1991;51:2979-84. 10. Bartek I, Bartkowa J, Vojtesek B, Staskova Z, Lukas 1, Rejhar A, et al. Aberrant expression of the p53 oncoprotein is a common fearure of a wide spectrum of human malignancies. Oiicogene 1991;6:1699-703. 11. Lane DP. On the expression of p53 protein in human cancer. Mol Uiol Rep 1994; 19:23-9. 3. 2011 12. Ahlgren JD, Macdonald JS. Gastrointestinal oncology. Philadelphia: J B Lippincott, 1992; 151-93. Tong ‘I,Cancer Statistics 1993. LA 13. Boring CC. Squires ‘IS, Cancer J Clin 1993,43:7-26. 14. Moriguchi S,Maehara Y, Korenaga D, Sugimachi K. Ilayashi Y,Nose Y.Prediction of survival time after curative surgery for advanced gastric cancer. Eicr J Surg Oncol 1992; 18:28792. 15. Nakajima ‘I, Kishi M, Kajitani T. Improvement in treatment results of gastric cancer with surgery and chemotherapy experience of 9700 cases in the Cancer Institute Hospital, Tokyo. Semin Sirrg Oncol 1991;7:365-72. 16. Agboola 0. Adjuvant treatment in gastric cancer. C m w r Treat Reu 1994;20:217-40. 17. Martin HM,Filipe MI, Morris RW. Lane I)€’, Silvestre F. p53 expression and prognosis in gastric carcinoma. Inf 1 Cr7ncer 1992;50359-62. 18. IJchino S, Koguhi .M, Hirota T, Itabashi M, Saito T, Kobayashi M, et al. Iiigh incidence of nuclear accumulation of p53 protein in gastric cancer. Jpn J Cliff Onto/ 1992;22:225-31. 19. Kakeji Y, Korenaga I), Tsujitani S,Baba 11, Anai f I , Macihard Y, et al. Gastric cancer with p53 overexpression has high potential for metastasing to lymph nodes. U r 1 Cuncer 1993;67:589-93. 20. Jopaul BV, Hopwood D, Kewman EL, Qureshi S, Granr A, Ogston SA, et al. The prognostic significance of the acciimulation of p53 tumour-suppressor gene protein in gastric adenocarcinoma. Rr J Cuncer 1994;69:943-46. 21. .Motojima K, Furui J. Kohara N , Ito T, Kanematsu T. Expression of p53 in gastric carcinomas is not independently prognostic. Surgev 1994; 116:890-95. 22. Starzynska T, Broniley M, Chosh A, Stern PI.. Prognostic significance of p53 overexpression in gastric and colorectal carcinoma. W r J (,’ancer 1992;66:558-62. 23. Ilermanek P, Sobin LH. editors. UICC ‘I‘NM Classification of malignant tumors. 4 th. ed. 2nd rev. Berlin: SpringerVerlag, 1992. 24. Japanese Research Society f o r Gastric Cancer. The general rules for gastric cancer study in surgery and pathology. lprl J Surg 1981;11:127-39. 25. Lauren P. The two histological main types of gastric icarcinoma: diffuse and so-called intestinal-type carcinoma: an attempt at a histo-clinical classification. Acta Patho/ Microbiol Scard 1965;64:31-49. 26. Peto R, Pike MC, Armitage f’, Hreslow NE, Cox DH, Howard SV. Design and analysis of randomised clinical trials rcquiring prolonged observation of each patient. I . Introduction and Design. Mr J Cancer 1976;34:585-612. 27. Peto R, Pike MC, Armitage P, 13reslow NE, Cox DH. Howard SV, et al. Design and analysis of randomised clinical trials requiring prolonged observation of each patient. 11. Analysis and examples. Br I Cancer 1977;3.5: 1-39. 28. Cox DR. Regression models and life tables. J H S f u t Soc H 1972;34:187-220. 29. Fielding I.P, Henson DE. Multiple prognostic factors and outcome analysis in patients with cancer. Cuncer 1993; 71:24‘L6-9. 30. Davidoff AM, I Ierndon JE, Glover NS. Kerns B-JM, Pence I(;, Iglehard ID, et al. Relation between p53 overexpression and established prognostic factors in breast cancer. Siirgery 1991; 110:259-64. 31. Hemvicos Y, ‘I’ominaga 0. Hammel P, Lauren-Puig P, Salmon RJ, Dutrillaux R, et al. Increased p53 protein content of colorectal turnours correlates with poor survival. Br J Cuncer 1992;66:7.58-64. 2012 CANCER May 15,1996 / Volume 77 / Number 10 32. Thor AD, Moore I1 DH, Edgerton SM. Kawasaki ES, Reihsaus E, Lynch HT, et al. Accumulation of p53 tumor suppressor gene protein: an independent marker of prognosis in breast cancers. J Natl Cancer Inst 1992;84:845-55. 33. Allred DC. Clark GM, Elledge R, Fuqua AW, Brown RW, Chamness GC, et al. Association of p53 protein expression with tumour cell proliferation rate and clinical outcome in node-negative breast cancer. J Natf Cancer Inst 1993;85:200-6. 34. Zeng B Z-S, Sarkis AS, Zhang Z-F, Klimstra DS, Charytonowicz E, Guillem JG, et al. p53 nuclear overexpression: an independent predictor of survival in lymph node-positive colorectal cancer patients. J Clin 0 1 7 ~1994; 0 ~ 12:2043-50. 35. Yonemura Y, Fushida S, Tsugawa K, Ninomiya I, Fonseca L. Y a m a m A, et al. Correlation of p53 expression and proliferative activity in gastric cancer. Anal Cell Parhol 1993;5:277-88. 36. Brito MJ, Williams GT, Thompson H, Filipe MI. Expression of p53 in early (Tl) gastric carcinoma and precancerous adjacent mucosa. Gur 1994;35:1697-700. 37. Fonseca L, Yoneniura Y, De Aretxabala X, Yamaguhi A, Miwa K, Miyazaki I. p53 detection as a prognostic factor in early gastric cancer. Omology 1994;51:485-90. 38. Oiwa S, Maehara Y, Ohno S, Sakaguchi Y, Ichiyoshi Y, Sugimachi K. Growth pattern and p53 overexpression in patients with early gastric cancer. Cancer 1995;75x1454-9.