Annals of Surgical Oncology 10:190-195 (2003)
© 2003 Society of Surgical Oncology
Expression of p53 and Proliferation Index as Prognostic Factors in Gastrointestinal Sarcomas
Heriberto Medina-Franco, MD,
Antonio Ramos-De la Medina, MD,
Ruben Cortes-Gonzalez, MD,
Jorge Baquera, MD,
Arturo Angeles-Angeles, MD,
Marshal M. Urist, MD,
Isam E. Eltoum, MD and
Martin J. Heslin, MD
From the Departments of Surgery (HM-F, AR-D, RC-G) and Pathology (JB), National Institute of Medical Sciences and Nutrition, "Salvador Zubiran," Mexico City, Mexico; and the Section of Surgical Oncology (MMU, MJH) and Department of Surgical Pathology (IEE), Division of General Surgery, The University of Alabama at Birmingham, Birmingham, Alabama.
Correspondence: Address correspondence and reprint requests to: Heriberto Medina-Franco, MD, Department of Surgery, National Institute of Medical Sciences and Nutrition, "Salvador Zubiran," Vasco de Quiroga 15, Tlalpan, Mexico City 14000, Mexico; Fax: 5573-9321; E-mail: herimd{at}hotmail.com
 |
ABSTRACT
|
|---|
Background: Sarcomas arising in the gastrointestinal (GI) tract are rare tumors. Molecular markers could be associated with prognosis in these types of tumors.
Methods: We performed a retrospective analysis of adult patients with sarcomas arising in the GI tract at the National Institute of Medical Sciences in Mexico City and the University of Alabama at Birmingham Hospital. All histological types were included. Patient, tumor, and treatment factors were analyzed, with overall survival as the main outcome variable. Expression of p53 and cellular proliferation antigen Ki-67 was also analyzed. Statistical analysis was performed by log-rank test and Cox regression. Significance was defined as P < .05.
Results: Forty-seven patients were analyzed. The median patient age was 53 years (range, 16–82 years). Twenty-five patients (53%) were women. The stomach was the most common site of presentation. The mean tumor size was 14 cm (2–45 cm). A complete resection was achieved in 40 patients. With a median follow-up of 30 months, the actuarial 3-year survival was 68%. Univariate analysis identified overexpression of p53 and Ki-67, high tumor grade, tumor size >10 cm, and incomplete resection as significant negative prognostic factors. Hispanic race and good performance status were significantly associated with prolonged survival. On multivariate analysis, overexpression of p53 was the only independent negative prognostic factor.
Conclusions: Overexpression of p53 is the strongest predictor of poor prognosis in patients with sarcomas of the GI tract.
Key Words: Gastrointestinal sarcoma • p53 • Proliferation index • Prognostic factors
|
INTRODUCTION
|
|---|
Sarcomas arising in the gastrointestinal (GI) tract are uncommon tumors, accounting for only .1% to 3% of all GI malignancies and approximately 10% of all sarcomas.1 Clinicopathologic determinants of survival have been conflicting in some aspects, despite several series of smooth muscle tumors of the digestive tract.2–7 Recently, there has been increasing interest in this type of tumor because they are responsive to molecularly targeted therapy.8,9 In addition, recent studies with molecular markers have suggested that alterations in p53 activity and the overexpression of Ki-67 as a proliferation marker are associated with prognosis in high-grade extremity soft tissue sarcomas.10 Most of the published reports of GI sarcomas span long periods of time, during which there may have been many changes in diagnostic and therapeutic approaches, and some have included benign mesenchymal tumors.11,12
The purpose of this review was to analyze the clinical presentation, management, histological variables, and factors associated with prognosis, with special focus on molecular markers, of primary GI sarcomas at the University Hospital, School of Medicine of the University of Alabama at Birmingham (UAB), Birmingham, AL, and the National Institute of Medical Sciences and Nutrition (NIMSN) in Mexico City, Mexico.
|
PATIENTS AND METHODS
|
|---|
Adult patients with a diagnosis of sarcoma arising in the GI tract at both institutions between January 1990 and December 2000 were included. We did not include any tumor classified as a benign stromal tumor or sarcomas of retroperitoneal origin. Charts and histopathologic slides were reviewed to determine demographic data, clinical presentation, histological type and grade of the tumor, extent of surgical resection, operative morbidity and mortality rates, the use of adjuvant therapy, patterns of recurrence, and overall survival data.
All available histological material was reviewed by a single pathologist at each institution (I.E.E for UAB and J.B. for the NIMSN). Original hematoxylin and eosin-stained slides were reviewed to confirm the presence of the tumor. Histological type was recorded as originally reported, except for four cases of ganglionic autonomic nerve tumors that were reclassified as gastrointestinal stromal tumors. Paraffin-embedded tissue blocks, when available, were retrieved and matched to confirm the same tissue block as was used in the original hematoxylin and eosin-stained slide. The grade of the tumor was assigned according to size, mitoses, tumor necrosis, nuclear pleomorphism, and cellularity.13
Immunostains for Ki-67 (clone MM1) and p53 (clone Bp53-11) were performed on formalin-fixed and paraffin-embedded material as described by the manufacturer (Ventana, Tucson, AZ). The number of positively staining nuclei was counted per 1000 tumor nuclei in the most intensely stained areas of the tumor. As other authors have suggested,10 we used nuclei with positive staining of 
20% to include as categorically positive.
For the purpose of this review, resection was considered complete if all gross disease was removed. Partial resection included removal of the bulk of the tumor but with gross disease remaining. Actuarial survival was measured by the Kaplan-Meier method beginning at the date of surgery. Univariate analysis to identify factors associated with survival was performed by the log-rank test, and multivariate analysis was performed with the Cox proportional hazards model. P < .05 was considered statistically significant.
|
RESULTS
|
|---|
During the study period, 49 adult patients with sarcomas arising in the GI tract were identified from the tumor registry from both institutions. One case each of rhabdomyosarcoma and peripheral neuroectodermal tumor was excluded because of their distinctive pathogenesis and clinical behavior, leaving 47 patients for analysis. Twenty-six patients (55.3%) were from the UAB database, and 21 (44.7%) were from the NIMSN. Demographic characteristics are listed in Table 1. There were 25 female and 22 male patients. Ages ranged from 16 to 82 years, with a median of 53 years. Most patients were Hispanic (45%). The stomach (51%) and small bowel (43%) were the most common sites of tumor origin. Other tumor locations are depicted in Fig. 1. Two patients had a history of other types of malignancies: a female patient with a history of breast cancer and a male patient with a history of bladder carcinoma.
Abdominal pain (67%), abdominal mass (57%), or both were the most common presenting symptoms. Other frequent complaints were GI bleeding (43%) and symptoms of obstruction (42%). Three quarters of patients presented with two or more symptoms, and only three patients were asymptomatic at diagnosis (the neoplasm was identified during a celiotomy for other reasons). The mean duration of symptoms was 7 months (range, 1–40 months). There were no patients with metastatic disease at presentation.
Forty-one patients had a preoperative computed tomography scan that was suggestive of GI sarcoma in 31 patients. Eight patients, all of them from UAB, underwent a preoperative biopsy. In three cases it was performed during GI endoscopy, and in five patients it was obtained percutaneously. In seven of these patients, there was a correct pathologic diagnosis. All patients were initially treated surgically. Forty patients (85%) underwent resection with curative intent; in other words, all tumor was removed with negative gross surgical margins. In 39 of these cases, microscopic margins were also reported free of tumor. Five patients had a debulking procedure to palliate symptoms, and two patients had only a tumor biopsy because of extensive and/or unresectable disease. Of 45 patients who underwent resection, adjacent organs were resected in 17 in addition to the primary tumor site. The most common organs resected were the pancreas (12 patients), spleen (6 patients), and colon (4 patients). Eleven patients had two or more organs resected. There was no operative (30-day) mortality.
Eleven patients received doxorubicin-based chemotherapy. In eight cases, it was administered in an adjuvant setting after gross tumor resection, and three patients received treatment in the presence of macroscopic disease. Three patients received radiotherapy, all of them in the presence of macroscopic disease.
There was an even distribution between gastrointestinal stromal tumors (n = 24) and leiomyosarcoma (n = 23), according to the original pathologic diagnosis. Twenty-three tumors were classified as high grade (49%) and 24 as low grade (51%). There were no regional nodes involved by tumor in any case. The tumor size was known in 45 of the 47 cases. The mean tumor size in these cases was 13.6 cm (range, 2–45 cm). Twenty-two tumors (49%) were 10 cm.
Immunohistochemistry for the expression of p53 or Ki-67 could be performed in 29 cases because paraffin-embedded material was not available for all cases. The molecular markers were considered positive in 13 cases (44.8%) each. In 10 cases (34.4%), both molecular markers were both positive. Ki-67 and p53 were more frequently expressed in high-grade (69%) than in low-grade (21%) tumors (P = .01) and in big (>10 cm; 77%) than in smaller (<10 cm) tumors (25%; P = .001). They were less commonly expressed in resected (30%) than in unresectable (100%) cases (P = .0001).
Of the 40 patients who were resected with curative intent, there were 12 recurrences: 2 local and 10 distant. There was no statistical difference in the possibility of recurrence according to histological diagnosis, but it was more common between patients with tumors positive for p53 and Ki-67. The most common site of distant failure was the liver (seven cases), followed by the peritoneum (three patients). The mean time to recurrence was 11 months (range, 4–30 months). Seventy-four percent of recurrences occurred in patients with high-grade tumors.
Overall actuarial 3- and 5-year survival was 68% and 57%, respectively. The median follow-up was 30 months (range, 1–112 months). The prognostic factors analyzed and their effect on survival are listed in Table 2. On multivariate analysis, the only factor associated with an adverse outcome was positivity for p53 (P = .04). The Kaplan-Meier curves for p53, tumor grade, and completeness of resection are presented in Figs. 2, 3, and 4, respectively.
|
DISCUSSION
|
|---|
GI sarcomas occur infrequently and account for only .1% to 3% of all GI malignancies and 10% of all sarcomas.1 Recently, they have been shown to be responsive to molecular targeted therapy. Most of these tumors are now known to be derived from the interstitial cells of Cajal, are characterized by expression of the proto-oncogene c-kit, and often harbor gain-of-function KIT mutations that lead to ligand-independent kinase activation. GI sarcomas are unresponsive to standard sarcoma chemotherapy, but in two recent phase II studies, STI571, which selectively inhibits BCR-ABL, KIT, and platelet-derived growth factor receptor tyrosine kinases, showed activity in patients with this type of neoplasm.8,9 Those reports have renewed interest in this kind of malignancy. In this article, we combined the experiences of two large referral centers in two countries so that we could analyze the biology of this tumor in different populations.
Of anatomical sites of the digestive tract, the stomach is the most common location of primary malignant stromal tumors: 38% to 65% of all GI sarcomas in most series3,5–7 and 51% in our series. Although somewhat less common in the small bowel (43% in our series), sarcomas represent the second most common small-bowel malignancy, with an incidence one third to one half that of adenocarcinoma.6 Other locations are less common and are represented by four cases of colorectal sarcomas and an isolated case of gallbladder rhabdomyosarcoma, which is extremely rare.14 This tumor and one case of rectal peripheral neuroectodermal tumor were excluded from analysis because they have distinctive pathogenesis and clinical behavior, and some authors do not consider those lesions as having a GI stromal origin.
The mean tumor diameter in our series was 14 cm (range, 2–45), which is bigger than that reported by McGrath et al.5 (11 cm) and Conlon et al.15 (9 cm). Ng et al.16 noted that 60% of GI sarcomas were >10 cm at presentation (49% in our series). There was almost an even distribution between high- and low-grade tumors in this series.
Consistent with other reports in the literature,5–7 this analysis demonstrated that completeness of resection, tumor grade, and size are significant prognostic factors on univariate analysis. Performance status was also statistically significantly related to prognosis. Race was not considered a prognostic factor in previous reports,7,15 but blacks and Hispanics have been found to have a worse prognosis in other neoplasms, such as breast cancer.17,18 Those differences have been attributed to socioeconomic status and a tendency to more advanced tumors in minorities.19 Strikingly, in this series, Hispanics were found to have a better prognosis than whites or blacks, with no difference in tumor grade, size, or completeness of resection between the populations of the two institutions included in this series (data not shown). At present, we do not have an explanation for this relationship.
Another factor analyzed was the positivity for Ki-67, which is an antigen expressed throughout most of the cell cycle and which is a good measure of dividing cells.20 Drobnjak et al.21 examined 174 fresh-frozen primary and recurrent soft tissue sarcomas from a variety of sites and correlated this with overall survival. In approximately 40% of the tumors, 20% of the cells expressed nuclear staining, with the higher percentage in synovial sarcomas (78%) and the lowest in liposarcomas/fibrosarcomas (27%). Similar to this study, a high proliferation index (20% nuclear expression) was associated with high-grade tumors and was correlated with poor outcome, but this did not retain independent prognostic significance when grade was taken into account. Choong et al.22 examined 182 paraffin-embedded primary extremity and superficial trunk soft tissue sarcomas for overexpression of Ki-67 (overexpression was defined as >10% nuclear staining) and demonstrated a significant correlation with size, grade, necrosis, vascular invasion, S-phase fraction, and metastasis. In this series, we were able to show a statistically significant relationship between grade and size and the expression of Ki-67, and, similar to the study of Choong et al., when subjected to multivariate analysis, overexpression of Ki-67 did not retain independent prognostic significance after other prognostic factors were taken into account.
Nuclear accumulation of the p53 protein indicates a longer half-life of the protein, suggests mutations in the DNA,23 and has been correlated with poor outcome in a number of tissue types, including soft tissue sarcoma.24 Most studies have been performed with fresh-frozen tissue and have revealed nuclear accumulation of p53 on the order of 23% to 33%.24–26 This study used paraffin-embedded tissue and revealed nuclear accumulation in 45% of the tumors analyzed. Although there was a relationship between overexpression of p53 and completeness of resection and grade, on multivariate analysis this factor retained its significance as a marker of poor outcome. In the study of Heslin et al.,10 p53 was not associated with outcome; however, only 9% of the tumors analyzed in this study showed overexpression of p53. Larger studies are necessary to confirm these results. At present, we are analyzing the relationship between this molecular marker and expression of c-kit.
In conclusion, this study has demonstrated that overexpression of p53 is an independent prognostic factor associated with an increased risk of recurrence and death in a group of patients with primary GI sarcomas. With clinical/radiological size, grade, Ki-67 index, and expression of p53 available, patients can be better stratified by risk, now that promising molecular targeted therapies are available.
|
Footnotes
|
|---|
Overexpression of p53 is the strongest predictor of poor prognosis in primary gastrointestinal sarcomas.
Received for publication March 26, 2002.
Accepted for publication December 12, 2002.
|
REFERENCES
|
|---|
- Medina-Franco H, Eltoum IE, Urist MM, Heslin MJ. Primary gastrointestinal sarcomas. Am Surg 2000; 66: 1–5.[Medline]
- Akwari OE, Dozois RR, Weiland LH, Beahrs OH. Leiomyosarcomas of the small and large bowel. Cancer 1978; 42: 1375–84.[CrossRef][Medline]
- Evans HL. Smooth muscle tumors of the gastrointestinal tract. Cancer 1985; 58: 2242–50.
- Zollinger RM, Sternfeld WC, Schreiber H. Primary neoplasms of the small intestine. Am J Surg 1986; 151: 654–8.[CrossRef][Medline]
- McGrath PC, Neifield JP, Lawrence W Jr, et al. Gastrointestinal sarcomas: analysis of prognostic factors. Ann Surg 1987; 206: 706–10.[Medline]
- Dougherty MJ, Compton C, Talbert M, Wood WC. Sarcomas of the gastrointestinal tract: separation into favorable and unfavorable prognostic groups by mitotic count. Ann Surg 1991; 214: 569–74.[Medline]
- DeMatteo RP, Lewis JJ, Leung D, Mudan SS, Woodruff JM, Brennan MF. Two hundred gastrointestinal stromal tumors. Recurrence patterns and prognostic factors for survival. Ann Surg 2000; 231: 51–8.[CrossRef][Medline]
- Blanke CD, von Mehren M, Joensuu H, et al. Evaluation of the safety and efficacy of an oral molecularly-targeted therapy, STI571, in patients with unresectable or metastatic gastrointestinal stromal tumors (GISTS) expressing c-kit (CD117). Proc Am Soc Clin Oncol 2001; 20: 2.
- Van Oosterom AT, Judson I, Verweij J, et al. STI571, an active drug in metastatic gastrointestinal stromal tumors (GIST), an EORTC phase I Study. Proc Am Soc Clin Oncol 2001; 20: 2.
- Heslin MJ, Cordon-Cardo C, Lewis JJ, Woodruff JM, Brennan MF. Ki-67 detected by MIB-1 predicts distant metastasis and tumor mortality in primary, high-grade extremity soft tissue sarcoma. Cancer 1998; 83: 490–7.[CrossRef][Medline]
- Sanders L, Silverman M, Rossi R, Braasch J, Munson L. Gastric smooth muscle tumors: diagnostic dilemmas and factors affecting outcome. World J Surg 1996; 20: 992–5.[CrossRef][Medline]
- Ludwig DJ, Traverso LW. Gut stomal tumors and their clinical behavior. Am J Surg 1997; 173: 390–4.[CrossRef][Medline]
- Lewin KJ, Riddel RH, Weinstein WM, eds. Gastrointestinal Pathology and Its Clinical Implications. New York: Igaku-Shoin Medical Publishers, 1992: 299–301.
- Al-Jaberi TM, Al-Masri N, Tubukhi A. Adult rhabdomyosarcoma of the gallbladder: case report and review of published works. Gut 1994; 35: 854–6.[Abstract/Free Full Text]
- Conlon KC, Casper ES, Brennan MF. Primary gastrointestinal sarcomas: analysis of prognostic factors. Ann Surg Oncol 1995; 2: 26–31.[Abstract]
- Ng EG, Pollock RE, Munsell MF, et al. Prognostic factors influencing survival in gastrointestinal leiomyosarcomas. Ann Surg 1992; 215: 68–77.[Medline]
- Daly MB, Clark GM, McGwire WM. Breast cancer prognosis in a mixed Caucasian-Hispanic population. J Natl Cancer Inst 1985; 74: 753–6.
- Elledge RM, Clark GM, Chamness GC, et al. Tumor biologic factors and breast cancer prognosis among white, Hispanic and black women in the United States. J Natl Cancer Inst 1994; 86: 705–9.[Abstract/Free Full Text]
- Gordon NH, Crowe JP, Brumberg DJ, et al. Socioeconomic factors and race in breast cancer recurrence and survival. Am J Epidemiol 1992; 135: 609–15.[Abstract/Free Full Text]
- Gerdes J. Ki-67 and other proliferation markers useful for immunohistological diagnostic and prognostic evaluations in human malignancies. Semin Cancer Biol 1990; 1: 199–206.[Medline]
- Drobnjak M, Latres E, Pollack D, et al. Prognostic implications of p53 nuclear overexpression and high proliferation index of Ki-67 in adult soft tissue sarcomas. J Natl Cancer Inst 1994; 86: 549–54.[Abstract/Free Full Text]
- Choong PF, Akerman M, Willen H, et al. Prognostic value of Ki-67 expression in 182 soft tissue sarcomas. Proliferation—a marker of metastasis? APMIS 1994; 102: 915–24.[Medline]
- Reich NC, Oren M, Levine AJ. Two distinct mechanisms regulate the levels of a cellular tumor antigen, p53. Mol Cell Biol 1983; 3: 2143–50.[Abstract/Free Full Text]
- Cordon-Cardo C, Lares E, Drobnjak M, et al. Molecular abnormalities of mdm2 and p53 genes in adult soft tissue sarcomas. Cancer Res 1994; 54: 794–9.[Abstract/Free Full Text]
- Leach FS, Tokino T, Meltzer P, et al. p53 mutation and MDM2 amplification in human soft tissue sarcomas. Cancer Res 1993; 53 (Suppl): 2231–4.[Abstract/Free Full Text]
- Andreassen A, Oyjord T, Hovig E, et al. p53 abnormalities in different subtypes of human sarcomas. Cancer Res 1993; 53: 468–71.[Abstract/Free Full Text]
TOP
ABSTRACT
INTRODUCTION
