Introduction
Gastrointestinal stromal tumor (GIST) is one of the most important groups of alimentary mesenchymal tumors. The tumors occur typically in the gastrointestinal tract from esophagus to anus [
1]. GISTs arise from interstitial cells of Cajal or the gastrointestinal pace-maker cells, and both of them share the characteristics of differentiating toward smooth muscle and neural cells [
2,
3]. It is reported that GISTs were associated with oncogenic activation of the KIT tyrosine kinase or its homolog platelet-derived growth factor receptor alpha (PDGFRA) [
4,
5]. Advancement in the understanding of GISTs pathogenesis has improved long-term survival of these patients. However, considerable controversies exist about the diagnosis, prognosis and outcomes of GISTs.
GISTs are diverse in biologic behaviors. It is difficult to predict the prognosis for an individual patient after surgical resection [
6]. The 5-year survival rate of low risk GISTs is approximately 95%, similar to the common population, while in high risk, the 5-year survival rate is less than 30% [
7]. National Institutes of Health (NIH) suggests tumor size and mitotic index are considered to be the most reliable prognostic indictors. However, debate has been raised about the accuracy of NIH criteria [
8]. This research was designed to assess the value of NIH criteria and identify other potential prognostic indictors that affect the prognostic course of patient with high risk GISTs.
Materials and methods
Patients’ characteristics
84 patients enrolled from January 1, 2005 to January 1, 2009 were analyzed after complete resection of the tumor. Clinicopathologic data including gender, age, date of surgery, tumor site, size, the patient’s disease condition and the time interval of tumor-associated death were collected. The study was approved by the institutional ethics board of Wuhan Union Hospital. The final diagnosis of GISTs was confirmed by histological and immunohistochemical analysis. Considering a part of GISTs were CD117-negative, tumors that fulfilled the clinical and morphological criteria but lacked CD117 expression were also included. We defined the moderate and high risk GISTs as malignant tumors, basing on the evidence that malignant behaviors belonged to the high or intermediate risk category and rarely were found in very-low or low group [
9,
10]. According to the above criteria, 84 patients were diagnosed as malignant GISTs.
Surgical approach
All patients underwent a radical abdominal exploration to make sure there were no metastatic diseases. Complete resections including the involved organ with an intact pseudocapsule were performed. Because no evidence showed lymph nodes resection could prolong overall survival [
11] and lymphadenectomy was not essential in surgical treatment [
12], lymphadenectomy was not performed routinely.
Survival and statistical analysis
Progression-free survival (PFS) was calculated from the date of surgery to the date of disease progression. Deaths from other reasons were considered as censored. The median follow-up time was 844 days, ranging from 52 to 2 145 days. Survival curves were calculated according to the Kaplan-Meier method and differences between curves were compared by log-rank test. SPSS 16.0 version was used to perform statistical analysis. A probability value less than 0.05 was considered as statistically significant.
Results
Patients’ characteristics
Four of 84 patients were excluded from our research for extensive abdominal metastases. Clinical records showed that the patients’ age ranged from 28 to 77 years (median 54 years). The age and gender distributions were summarized in Fig. 1. Symptoms of patients with GISTs were vague and non-specific. Clinical manifestation may be presented with mass-associated symptoms or anemia. The most common complaint was abdominal discomfort (43.8%), followed by gastrointestinal bleeding (37.5%), hematemesis, and occasionally chronic anemia. There were 10 patients (12.5%) who palpated their own mass externally and 5 asymptomatic cases (6.2%) who were detected incidentally for other reasons.
Tumor features and treatment
32.5% (26 of 80 cases) was involved in stomach and 38.75% (31 of 80) in small intestine (including duodenum, jejunum and ileum). There were 4 patients whose tumor occurred in extra-gastrointestinal sites (two cases in mesentery, one in omentum majus and one in pulmonary artery). The malignant tumors ranged in size from 2 cm to 30 cm (median 10 cm). Presence of necrosis was observed in 35 cases (43.8%). Mitotic index varied between 0 to 55 mitoses per 50 high power field (HPF), and the cases negative for CD117 were 4 (5.0%). Complete resections without routine lymphadenectomy were performed in all patients. The clinical characteristics of enrolled patients were summarized in Table 1.
Prognostic factors evaluation
At the end of the study, 39 patients were alive and 4 patients occurred recurrence or metastasis. The median follow-up time was 844 days (52-2 145 days) after complete resection of primary tumor. Several clinic data variables were analyzed to identify the prognostic factors of malignant GISTs. They were tumor size, site, mitotic index and necrosis. Fig. 2 showed that tumors with intermediate risk had more favorable prognosis than that with high risk. Their 3-year survival rates were 65.3% and 41.3%, respectively (P<0.001). Anatomic location was also a prognostic factor. As shown in Fig. 3, the 3-year survival rate of tumors in stomach is 66.1% while tumors in small intestine was 51.4% (P = 0.017). Gastric GISTs tend to behave in a less aggressive mode than small intestinal GISTs. As NIH criteria suggested, tumor size and mitotic index had a correlation with progression free survival. The 3-year survival rates for patients with tumor size≤10 cm and>10 cm were 62.3% and 41.8%, respectively (P = 0.002, Fig. 4). GISTs with a diameter>10 cm had an increased probability of malignant behavior. Tumors with mitotic index≤5/50 HPF had a higher 3-year survival rate than tumors with mitotic index>5/50 HPF (67.1% versus 40.7%, P = 0.005, Fig. 5). The presence of necrosis was directly associated with the poor prognosis. The 3-year survival rates for presence and absence of necrosis were 50.8% and 64.8% (P = 0.008, Fig. 6).
Discussion
Over the past several years, many efforts have been made on the treatment of GISTs. Advances in molecular genetics analysis deepen our understanding on the pathogenesis of GISTs, and the process of tumors risk category is also evolving [
13]. Surgery is still the primary approach to the GISTs. For most very low and low risk patients, complete resection is curative, and these patients have an excellent prognosis. The 5-year survival rate of low risk GISTs is approximately 95%, similar to the normal population. However, prognosis for intermediate and high risk group is poor. It is desirable to define the factors that affect biologic behavior of malignant GISTs.
The study indicated that classification from intermediate risk to high risk tumors correlated well with survival, which was in agreement with previous report [
14,
15]. Biologic behavior of intermediate risk GISTs had a better prognosis than high risk GISTs. Our study also confirmed that a mitotic index of more than 5/50 HPF was a poor prognostic indicator. Tumor size was also associated with GISTs behavior. There was a decrease in overall survival with increasing size. Patients with tumors larger than 10 cm had a poor overall survival than those whose tumors were 10 cm or less. Even though tumors with low mitotic index had a favorable prognosis, it was not certain that low count suggested benign prognostic course [
16]. The most frequent site of origin was small intestine, followed by stomach and rectum. The clinical course of GISTs seemed to be site-dependent. Gastric GISTs tended to behave in a less aggressive fashion than small intestinal GISTs. The result was consistent with other reports [
17,
18]. However, the mechanism was unclear. Haller
et al. observed that expressions of platelet-derived growth factor receptor alpha (PDGFRA) and KIT were different between stomach and small intestine [
19]. Expression of PDGFRA was higher and expression of KIT was lower in gastric GISTs when compared with GISTs in small intestine. Higher expression of PDGFRA may be the reason for prognostic difference. Gene analysis is important in GISTs treatment, because it can predict tumors behavior as well as sensitiveness to adjuvant imatinib mesylate [
20-
22]. It is reported that effectiveness of imatinib was related to KIT mutation status [
3,
23]. Tumors with lower expression of KIT are likely resistant to imatinb treatment. Intra tumoral necrosis was observed in 45 cases (56.7%), and was significantly associated with aggressive behaviors compared to the tumors without necrosis. Besides tumor size and mitotic index, tumor location and necrosis were also predictive indicators. Some experts suggested these factors should be taken into consideration when we performed risk stratification of GISTs [
24,
25]. It is reported that the score proposed by Fletcher has better sensitivity and specificity to predict prognosis compared with modified NIH criteria [
26].
Surgery is still the major treatment for GISTs. Complete tumor resection with clear margins is required [
27]. To avoid tumor rupture and risk of intra-abdominal dissemination, the tumor should be handled with care. As for metastatic or invasive GISTs, if possible, surgical resection is still recommended [
28,
29]. Imatinib is the first line treatment for recurrent and unresectable tumors. Combining surgery and imatinib therapy may reduce recurrence and decrease the risk of disease progression in patients [
30]. GISTs recurrence can occur late after surgical resection, so a routine follow-up is necessary. For very low or low risk GISTs, even though relapse is rare in these groups, patients still should be informed the likelihood of recurrence. Follow-up with computed tomography imaging of abdomen is recommended based on every year. For patients with malignant GISTs (intermediate and high risk GISTs), follow-up is more important, which should be taken on every 3-6 months for at least 5 years [
31].
In conclusion, besides tumors size and mitotic index, tumor location and necrosis also affect long-term survival of patients with malignant GISTs. Physicians should be familiar with these prognostic factors, and establish personalized therapy. Gene mutation analysis is becoming more and more important in treatment, which can provide information about not only diagnosis, but also the tumor’s response to imatinib.
Higher Education Press and Springer-Verlag Berlin Heidelberg
PDF
(141KB)
