|Year : 2015 | Volume
| Issue : 3 | Page : 351-357
Evaluation of biologic potential and risk stratification for predicting disease-free survival after resection of primary gastrointestinal stromal tumor: A multivariate clinicopathological study
M Vij1, V Agrawal2, A Kumar3, R Pandey2
1 Department of Pathology, Global Health City, Chennai, Tamil Nadu, India
2 Department of Pathology, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India
3 Department of Surgical Gastroenterology, Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, Uttar Pradesh, India
|Date of Web Publication||18-Feb-2016|
Department of Pathology, Global Health City, Chennai, Tamil Nadu
Source of Support: None, Conflict of Interest: None
Background And Objectives: Gastrointestinal stromal tumor (GIST) is mesenchymal neoplasms of the gastrointestinal tract, which express CD117, a c-kit proto-oncogene protein and show gain of function mutation of c-kit gene. Apart from the presence of metastasis, the criteria to differentiate benign and malignant GISTs are not well-defined. Although a variety of prognostic factors have been investigated, no method has yet proven sufficient to enable reliable determination of malignancy in all cases. This study was planned to risk stratify the GIST cases with respect to the various clinicopathological features and to identify prognostic factors in GIST. Materials And Methods: n histological and immunohistochemical analysis, 121 cases of GIST were identified. MIB-1 (Ki-67) labeling index (LI) was performed in 60 cases. Follow-up data was available for 93 patients. A P < 0.05 was taken as significant. Results: Larger tumor size, high mitotic activity and Ki-67 LI of >10% were identified as significant predictors of disease-free survival in univariate analysis (P < 0.0001). Other factors of statistically significant value were a high cellularity (P < 0.0027), nuclear pleomorphism (P = 0.0002), epithelioid cell type (P = 0.0098), presence of tumor necrosis (P < 0.01), presence of skeinoid fibers (P = 0.042), S-100 negativity (P = 0.025). Extra-gastrointestinal GIST and metastasis were more frequently associated with progressive disease (PD) as compared with GIST (P < 0.0004), (P < 0.0001). On multivariate analysis size (P = 0.0025), Ki-67 labeling index (P = 0.0186) and mitotic count (P = 0.0375) emerged as independent prognostic predictors of PD. Conclusion: This study suggests that GIST in Asian population may have a different phenotype with some predilection to nodal metastasis. Of all the features studied, tumor size and mitotic index are the best prognosticators in GIST with the addition of Ki-67 LI, wherever available.
Keywords: CD117, gastrointestinal stromal tumors, gastrointestinal tract, immunophenotyping, prognosis, survival
|How to cite this article:|
Vij M, Agrawal V, Kumar A, Pandey R. Evaluation of biologic potential and risk stratification for predicting disease-free survival after resection of primary gastrointestinal stromal tumor: A multivariate clinicopathological study. Indian J Cancer 2015;52:351-7
|How to cite this URL:|
Vij M, Agrawal V, Kumar A, Pandey R. Evaluation of biologic potential and risk stratification for predicting disease-free survival after resection of primary gastrointestinal stromal tumor: A multivariate clinicopathological study. Indian J Cancer [serial online] 2015 [cited 2019 Jun 26];52:351-7. Available from: http://www.indianjcancer.com/text.asp?2015/52/3/351/176689
| » Introduction|| |
Primary mesenchymal tumors of the intestinal tract are a heterogeneous group of tumors. Traditionally, the majority of these tumors were thought to be derived from smooth muscle cells. It has now been realized that most of these tumors are not smooth muscle or nerve sheath tumors, but represent a distinct clinicopathological entity termed as gastrointestinal stromal tumor (GIST). It is currently believed that GIST is a specific mesenchymal neoplasm and the term is used to refer to those mesenchymal neoplasms of the gastrointestinal tract (GIT), which express CD117, a c-kit proto-oncogene protein and show gain of function mutation of c-kit gene that encodes a growth factor receptor with tyrosine kinase activity., The introduction of a new targeted treatment for GIST in the form of imatinib mesylate, a receptor tyrosine kinase inhibitor, has further validated this entity.,
The criteria to differentiate between benign and malignant GISTs have been sought, analyzed and disputed for many years. Although a variety of prognostic factors have been investigated, no method or combination of methods has yet proven sufficient to enable reliable determination of malignancy in all cases. Many of the studies had reported that a tumor size of more than 5 cm and a mitotic count of >5/50 high-power fields (HPF) has high tumor-related mortality. Miettinen and Lasota have suggested guidelines for evaluating biological potential for GIST at various sites. Positivity for Ki67 analogs has been shown as the potential prognostic indicator of GISTs, with prognosis worsening with increasing positivity in some studies., This study was planned to characterize the mesenchymal tumors of the GIT by using morphological and immunohistochemical features into GIST and non-GIST tumors, to risk stratify the cases with respect to the various clinicopathological features and proliferation markers (Ki-67/MIB-1 labeling index (LI)) and to identify prognostic factors in GIST.
| » Materials and Methods|| |
From 1st January 1989 to 31st October 2007, all evaluable cases of mesenchymal tumors of the GIT, retroperitoneum and mesentery were retrieved from the database files of the department of pathology. Clinical and pathological data were retrieved from hospital information system, database files from Department of Surgical Gastroenterology, records of the Department of Pathology, post-operative follow-up and letters and telephonic conversation as and when required. A total of 133 cases of mesenchymal tumors of GIT were diagnosed during this period and all were included in this study. This included one hundred and six (106) retrospective cases and twenty six (26) prospective cases. One hundred twenty three (1) were gastrointestinal and thirteen (13) were extra-gastrointestinal. The material comprised of formalin-fixed paraffin embedded tissue blocks and tissue sections. Additional sections of 3-5 micron was cut and stained with hematoxylin and eosin as required. Immunohistochemical markers included CD117, CD34, smooth muscle actin (SMA), desmin, S100, vimentin, Ki67. For Ki67 proliferation index, Ki67 staining was carried out. For this 1000 cells were counted under HPF. Cells with nuclei stained were taken as positive. Positive cells percentage were calculated and were divided into three groups as 1%, 1-10% and >10%. The comparison of groups for continuous variables and counts was performed using Mann-Whitney U-test. The association between categorical variables was assessed using Pearson's Chi-square and Fisher's exact test. The correlation between pairs of variables was tested using Pearson's bivariate analysis. Survival analysis was performed using Kaplan-Meier analysis and log-rank test was used to compare the differences in survival between groups. Multivariate analysis for survival was performed using Cox regression analysis. Statistical analyses were performed using Microsoft Excel © (Microsoft, Redmond, United States of America) and Statistical Package for Social Sciences (SPSS © IBM SPSS, Inc. Chicago, Illinois, United States of America). A probability of <0.05 was considered statistically significant.
| » Observations and Results|| |
A total of 133 mesenchymal tumors of the GIT and stromal tumors of pancreas, mesentery and retroperitoneum, were received at the Department of Pathology, from 1st January 1989 to 31st July 2007. The age of patients was known in 132 patients and varied from 8 to 83 years (mean 49.76, median 48). There were 94 men with male female ratio of 2.4:1. The tumors were reclassified on the basis of morphological features and immunohistochemical findings into 121 GIST and extra GISTs and 12 other mesenchymal tumors including leiomyoma, n = 6, leiomyosarcoma, n = 2, schwannoma, n = 2), intestinal fibromatosis (n = 1) and inflammatory myofibroblastic tumor (n = 1).
GISTs included 83 men and 38 women with male female ratio of 2.15:1. Patient age range was 15-83 years (mean 50.42, median 50). The most common presentation of GIST was gastrointestinal bleeding. Gross dissemination, including local infiltration and/or metastasis at the time of surgery was noted in 26 patients. The tumor size varied from 0.3 cm in an incidentally detected tumor to a maximum of 35 cm in a gastric tumor. The mean tumor size was 10.71 cm and the median was 9 cm. The most common gross appearance of gastrointestinal tumors (N = 106) was described as exophytic serosal growth [Figure 1]a and [Figure 1]b or growth bulging out from muscularis propria (N = 60, 56.60%).
|Figure 1: (a) Gross appearance of a large cystic gastric gastrointestinal stromal tumor with hemorrhagic areas; (b) Gross appearance of a gastric GIST with dumbbell appearance; (c) Spindle cell GIST with juxtanuclear vacuolization; (d) Nested or paraganglioma-like growth pattern in low grade epithelioid GIST|
Click here to view
The cellularity of the majority of tumors was subjectively assessed as moderate (N = 67). The majority (N = 78, 64.46%) of tumors were classified as spindle cell type [Figure 1]c while 20 (16.52%) were classified as epithelioid type [Figure 1]d and 23 (19%) as mixed cell type. The most common growth pattern was fascicular 112/121 (92.56%) [Figure 2]a. Other patterns were diffuse (39.66%), palisading (23.96%) [Figure 2]b, nested (13.22%) and storiform (13.22%). Myxoid areas were also identified [Figure 2]c, the nuclear pleomorphism was recorded as moderate in (N = 54, 44.62%) and marked in 22 (18.18%) of tumors [Figure 2]d. Coagulative necrosis was present in 56.19% (68/121) tumor and comprised >1 low-power field in most of the tumors (32/68). Skeinoid fibers were seen in six small intestinal tumors [Figure 3]a. Mucosal invasion was identified in 28 tumors [Figure 3]b. The mitotic count ranged between 0 and 65/5 mm 2 and the majority of tumors (N = 80) had ≤5 mitotic figures per 5 mm 2. Categorization of GIST of different sites into various risk groups was performed as suggested recently by Miettinen and Lasota.
|Figure 2: (a) Fascicles of spindle cell gastrointestinal stromal tumor; (b) Prominent nuclear palisading in spindle cell GIST; (c) Tumor cells in bundles separated by myxoid stroma; (d) Marked nuclear atypia with bizarre cells in high grade GIST|
Click here to view
|Figure 3: (a) Skeinoid fibers observed in small intestinal gastrointestinal stromal tumor; (b) Mucosal invasion in GIST with tumor cells between the glandular elements; (c) Strong Golgi zone immunostaining of CD117 with weak cytoplasmic staining in GIST; (d) High Ki-67/MIB-1 labeling of nuclei in a case of GIST with progressive disease|
Click here to view
CD117 immunopositivity was seen in 114 (94.21%) tumors with strong intensity of positivity in the majority of tumors [Figure 3]c. The present study showed CD34 immunopositivity in 59.5% (72/121) GIST. SMA immunopositivity was seen in 48 tumors (39.67%). S100 immunopositivity was seen in 44 tumors and showed moderate to strong cytoplasmic and nuclear positivity. Immunopositivity for desmin was recorded in five tumors. Vimentin immunostaining was performed in 59/121 cases. Immunohistochemical estimation of Ki-67/MIB-1 expression was performed in 60 tumors. The range of MIB-1 LI was 0.2-31, mean and median being 6.41 and 3.6, respectively [Figure 3]d.
The follow-up data was available for 93 patients. The overall range of follow-up was 0-127 months. The mean follow-up was 26.53 months and median follow-up was 15 months. The follow-up revealed evidence of progressive disease (PD) in 49 patients (52.58%, mean follow-up 19.67 months, median 12 months) in the form of tumor recurrence, development of metastasis or dissemination, which was diagnosed on radiological studies or on abdominal exploration and histological confirmation. Five patients were known to have died of disease during follow-up. Three of these had gastric tumors, one had a jejunal tumor and one had a mesenteric tumor. Three had disseminated disease at presentation and the remainder developed disseminated disease during follow-up. PD was identified in 24 patients at first presentation. There was no evidence of PD in 44 patients (47.31%) during follow-up (mean follow-up 34 months, median 17.5 months). The evidence of PD was found in 24 patients with gastric tumors, 14 with small intestinal tumors, one with colorectal tumor and 10 with extra-gastrointestinal (pancreatic, mesenteric and retroperitoneal) tumors. The proportion of patients developing PD was significantly higher in extra-gastrointestinal tumors as compared with gastric and intestinal tumors (P = 0.03). The most common site for metastasis/dissemination was liver, followed by lymph nodes and pancreas. The overall mean disease-free survival (DFS) was 52.33 months. The overall range of DFS was 0-127 months and the median DFS was 27 months. The overall 1-year DFS was 64.5% and the overall 5-year DFS was 33.8%.
Although the overall DFS was slightly longer in female patients (mean DFS, 64.85 months) as compared with males (mean DFS, 47.69 months), this difference was not statistically significant (P = 0.15). There was no significant difference in DFS with respect to age of patients or site of tumor within the GIT. There was no significant difference in DFS in gastric GIST (mean, 55.99 months) as compared with intestinal GIST (mean, 61.16 months, (P = 0.75). The DFS in gastrointestinal GIST was significantly better as compared with DFS in extra-gastrointestinal GIST (EGIST) [P = 0.0004, [Figure 4]a. The mean DFS in gastrointestinal GIST was 58.64 months and mean DFS in EGIST was 10.10 months. The tumor size was found to be an important predictor of survival [P < 0.00001, [Figure 4]b. The mean DFS of patients with tumor size <5 cm was 54 months, while, the mean DFS for tumors >10 cm was 12.82 months.
|Figure 4: (a) Comparison of cumulative disease-free survival in gastrointestinal stromal tumor and extra-gastrointestinal GIST; (b) Kaplan Meier plots for DFS for GIST grouped according to size; (c) Cumulative DFS in GIST grouped according to cellularity; (d) Cumulative DFS in GIST grouped according to mitotic count per 5 mm2|
Click here to view
There was a significantly higher frequency of PD in patients with tumors displaying high cellularity [P = 0.0027, [Figure 4]c, nuclear pleomorphism (P = 0.0002), epithelioid cell type (P = 0.0098) and tumor necrosis (P < 0.01). The mean DFS in patients with nuclear pleomorphism was 39.49 months while that in tumors without significant nuclear pleomorphism was 89.82 months. The mean DFS of GIST of spindle cell type was 59.38 months while the mean DFS patients having GIST with prominent epithelioid features was 12.85 months. For tumors with a mixed cell type, the mean DFS was 56.51 months. The mean DFS for tumors without necrosis was 71.55 and median survival time 62.00. The mean DFS for tumors with necrosis was 39.81 and the median survival time was 15.00. The increasing amount of necrosis was also associated with a worse outcome with a linear trend (P = 0.0039). The presence of skeinoid fibers was associated with a significantly lower incidence of PD in patients with small intestinal GIST (P = 0.042). The presence or absence of mucosal invasion (P = 0.7350) and mucosal ulceration (P = 0.7426) as well as other histological findings such as hyalinization, myxoid degeneration and different growth patterns were not associated with significantly different incidence of PD. Mitotic activity proved to be a significant factor for predicting DFS in patients of GIST [P < 0.0001, [Figure 4]d. The mean DFS for patients with a mitotic count of <5/5 mm 2 was 80.29 months while the mean DFS for patients with mitotic count of >5/5 mm 2 was 21.80 months.
There was no significant difference in the incidence of PD between patients having tumors with or without CD117, CD34, SMA and desmin positivity or between patients with tumors displaying different patterns of positivity. However, S100 positivity proved to be significantly associated with better DFS (P = 0.025) in patients with GIST. Ki67 LI >10% emerged as a powerful predictor of reduced DFS (P < 0.0001). The mean DFS in patients with tumors having MIB-1 LI of more than 10% was 8.96 months while that for tumors having MIB-1 LI of less than 10% was 75.88 months.
We assessed the patients in the present study according to the current risk stratification according to Miettinen and Lasota. There was statistically significant higher evidence of PD in patient with a high risk groups (P < 0.0001). No metastasis was seen in risk groups 1, 2 and 5. Eight of 33 patients in the risk group 3a developed metastasis and one patient each in the risk groups 3b and 6a developed metastasis while 14 of 27 patients in group 6b developed metastasis. The data in the present study showed no significant difference between group 1 and 2, between group 3b and 5 and 6a and 6b. For the patients included in the present study, we combined patients from group 1 and 2 and group 3b and 5 and 6a and 6b. The DFS was significantly different in these derived risk groups (P < 0.0001). The mean DFSs in proposed risk groups ranged from 54 months in risk group 1-8.83 months in risk group 4. Metastasis was significantly associated with PD (P < 0.0001). Peritoneal metastasis was significantly associated with poorer outcome as compared to the liver or nodal metastasis (P < 0.0001).
Cox proportional hazard model was used to identify prognostic factors of independent significance. As MIB-1 LI was not available in all the tumors, we analyzed all the factors except MIB-1 LI that were found to be significant in univariate analysis by Kaplan-Meier analysis and log-rank test. Tumor size, mitotic index per 5 mm 2 and nuclear pleomorphism emerged as independent prognostic variables. However, when only those patients were included in the model where MIB-1 LI was available, tumor size and mitotic count retained their significance, but nuclear pleomorphism was replaced by MIB-1 LI as another parameter of independent prognostic significance.
| » Discussion|| |
GIST is a relatively rare neoplasm representing ~0.1-3% of all primary tumors of the GIT. We found 121 cases of GISTs of GIT, retroperitoneum, mesentery and pancreas out of 133 mesenchymal tumors during last 19 years. 108 (89.3%) GIST were located in the GIT and 13 (10.7%) in the extra-gastrointestinal sites. The median follow-up in this study was 15 months (mean: 26.53 months, range: 0-127 months). The mean recurrence free survival was 52.3 months (median: 27 months) and the median disease free survival after resection of primary GIST was 37 months. This is lower than most western studies.,,, The lower survival in our study is likely to be due to larger tumor size and late presentation of patients. Referral bias may also be an important factor as ours is a tertiary care referral hospital.
In the present study, EGIST were more frequently associated with PD as compared with GIST of tubular GIT (P < 0.0004). This is comparable with the study by Emory et al. who reported that peritoneal tumors had a worse prognosis compared with gastrointestinal GIST. Similar observations were made by Ahmed et al. Two important reasons for EGIST to have a poorer prognosis may be their larger size and late detection. A recent study of a small number of omental (nine cases) and mesenteric (seven cases) EGIST seemed to show a better prognosis for the former group. However, this study did not stratify these tumors for other parameters that arguably affect outcome in GIST. It has been suggested that tumors of the small intestine have a more aggressive course than tumors of the stomach. In the current study, we found no significant difference in DFS between gastric and small intestinal GIST (P < 0.7494) as reported in other studies.,
We observed a statistically significant recurrence free survival in tumors up to 5 cm in size compared with tumor size >5 cm to ≤10 cm and tumor size >10 cm (P < 0.0001). There was a linear trend in the decrease in recurrence free survival as the size of the primary tumor increased (P < 0.0001). This suggests an increasing risk of PD with increase in size. DeMatteo et al. have reported that tumor size was a statistically significant predictor of survival. Singer et al. reported that tumor size was statistically significant for an adverse outcome (P < 0.03). A recent study has found size to be of only borderline significance, which is in contrast to majority of other studies.
Tworek et al. have reported that cellularity was the most robust predictor of metastatic outcome in the multivariate analysis. Similarly, Miettinen et al. have reported that hypercellular spindle and epithelioid tumors had a worse outcome than sclerosing tumors. 32.83% (22/67) gastric GIST were of high cellularity. On follow-up, 12 of these showed PD. Only seven (10.44%) were of low cellularity, follow-up available in five of these patients showed no evidence of PD. This is comparable with a study by Trupiano et al. Cellularity has also been reported to be an important predictor of survival in small intestinal tumors., 27.8% of small intestine GIST in the present study showed high cellularity, which is slightly lower than that reported in literature., Follow-up was available in 7 of these patients and 6 showed evidence of PD. Reith et al. have showed 43.75% extra-gastrointestinal tumors to be of high cellularity and reported cellularity to be an independent prognostic factor in univariate analysis.
In the current study, epithelioid cell tumors were more often associated with PD followed by mixed tumors, while spindle cell tumors were least often associated with PD. Similar findings were reported by Rutkowski et al. and Wong et al., This finding however, is in contrast to many other major studies that have found epithelioid phenotype to be a favorable prognostic factor.,, In the largest study on gastric GIST, Miettinen et al. have reported that gastric GIST with platelet-derived growth factor receptor alpha (PDGFRA) mutation had epithelioid morphology and favorable prognosis, however it was not statistically significant. The series by Trupiano et al. reported no statistical significant difference between all the three cell types in adverse outcome group (P < 0.78). One reason for this difference could be a different population characteristics in this part of the world. Secondly, the present series included more sarcomatous epithelioid tumors, which show no difference in survival from spindle cell type. Thirdly, it may be surmised that the frequent PDGFRA mutation seen in low grade epithelioid GIST in western studies may be less frequent in the patients included in the present study. However, this hypothesis is difficult to substantiate in the absence of mutation studies.
We also found a linear trend between the amount of necrosis and PD (P = 0.0039) which was highly significant. Rudolph et al. and Iesalnieks et al. reported similar findings., Studies on gastric and small intestinal GIST have reported similar results., Nilsson et al. reported that degree of cellular pleomorphism was statistically significant with regard to overall survival. Diffuse atypia was a significant prognostic factor as compared with no atypia in the study by Miettinen et al. As described in most previous studies, we found mitotic count to be an important prognostic factor. GIST are usually divided in to two broad groups, based on mitotic index of up to 5 and more than 5/50 high-power microscopic fields. Since the areas of HPF vary in different microscopes, the mitotic index reported in various studies is not comparable as only an occasional study has specified the exact area, in which the mitotic figures were counted. However, it is now generally accepted that older denominator of 50 HPF should be replaced by 5 mm 2 for GIST. Nearly, all the studies have found a significant association of high mitotic count with adverse outcome.,,,, However, the reported mitotic index has varied widely in previous studies.,, We also noted significant correlation between mitotic count and tumor size (P < 0.001).
Mucosal invasion is usually significantly related to the prognosis. Miettinen et al. have reported it exclusively in patients with sarcomatous gastric tumors, a great majority of whom died of disease. Trupiano et al. have also reported that mucosal invasion predominated in the adverse outcome group in their study. Mucosal invasion in small bowel GIST usually been found to be associated with an adverse outcome. The presence of skeinoid fibers has consistently been reported with mitotically inactive tumors as a prognostically favorable factor.,, All six cases with skeinoid fibers showing no evidence of PD in the present study.
Immunohistochemical neural differentiation was significantly associated with better disease survival (P = 0.0248) in the present study. Rudolph et al. have also reported that neural differentiation was associated with slightly improved survival and lower incidence of metastasis. Wong et al. in a study of 108 gastric GIST found three cases to be positive for S100 and all were alive with a median follow-up of 35 months. We found no significant association of PD with CD34, SMA or desmin immunoreactivity. However, SMA and desmin positivity have been found to be prognostically favorable factors in a study by Miettinen et al. The importance of Ki-67 LI has been noted by some previous studies. However, different studies have used different cut-offs.,, Surprisingly, in three separate studies, indices below these very different cut off values of KI67 index resulted in remarkably similar survival curves.,, There was more variation in survival curves for tumors with KI67 index above these cut off values. Difference in patient cohorts and in definition of GIST may some of the contributing factors that cause this variability. There is some debate whether Ki-67 LI assessment is more practical and reliable than mitotic count. Although the former is an immunohistochemical technique that is costlier and may be subject to inter-laboratory variability, our findings support the application of Ki-67 immunohistochemistry in GIST along with careful assessment of mitosis/5 mm 2 as both these parameters supplement each other. One important observation by Yasui et al. was demonstration of inter-tumoral Ki-67 LI heterogeneity in gastric GIST. Since mitotic counts can also have intra-tumoral heterogeneity; assessment of both parameters may help to offset the heterogeneity in at least some tumors.
We found significant difference in survival between various prognostic risk groups as defined by Miettinen et al., with higher risk groups associated with higher incidence of PD (P < 0.0001)., However, there were no patients in group 4 in our study. Tumors belonging to this group are likely to be very rare as mitotically active tumors usually grow over the size limit of 2 cm before being detected. Significant difference in recurrence free survival was not observed between group 1 and 2 and between group 3b and 5. Therefore, for the purpose of this study we combined these risk groups and a created a simpler 4-tier classification [Table 1] and found them to be highly significant (P < 0.0001). However, as the present study comprises of only 121 cases with available follow-up in only 93 cases, this classification needs to be validated in a larger study and also by a large-population based study.
A finding of interest in the present study was the presence of lymph node metastasis in gastric GIST, which has not usually been reported in the western literature., The western literature in fact suggests not to do lymphadnectomy in these tumors., However, four gastric cases in the present study showed lymph node metastasis and two of these patients also had liver metastasis. Kim et al., in a study from Korea, also reported nodal metastasis in three patients. This is likely to be due to larger size or late presentation of tumors in these studies although it cannot be excluded that GIST in Asian population may have a different phenotype with some predilection to nodal metastasis.
On analysis of all the factors found to be significant in univariate analysis with Cox regression analysis with forward stepwise progression, size (P = 0.0025), Ki-67 LI (P = 0.0186) and mitotic count (P = 0.0375) emerged as independent prognostic predictors of PD. The importance of tumor size is consistently recorded in almost all previous studies.,,, The two largest Armed Forces Institute of Pathology studies confirmed tumor size to be the only independent prognostic factor., Rudolph et al. showed Ki-S5 score and atypical mitotic figures as the independent predictor of overall survival. On the other hand, Rutkowski et al. identified five independent factors that negatively influence DFS, including mitotic index of >5/50 HPF (P = 0.004), primary tumor size >5 cm (P = 0.001), male sex (P = 0.003), R1 resection/tumor rupture (P = 0.04) and primary extra gastric tumor location (P = 0.02). Iesalnieks et al. found that the presence of distant metastasis, MIB-1 index and mutation in kit exon 11 correlated significantly with the overall poorer survival. Steigen et al. showed that cases with more than 5 mitoses per 50 HPF had an increased expected mortality than of those with fewer mitoses. It appears that tumor size with mitotic index can still be considered the best prognosticators in GIST with possible addition of Ki-67 LI, wherever available.
To conclude, the present study showed that patients with high risk GIST had a high risk of recurrence/metastasis. However, despite these established prognostic factors, predicting clinical behavior of a GIST in a particular patient remains a challenge. It is accepted that all GISTs can exhibit malignant potential and none, except perhaps <1 cm tumors, can be labeled as benign. The optimal management of individuals who have undergone excision of high-risk tumors (based on proliferation index, tumor size and mitosis) remains under study. Based on the findings of the present study, a logical program would involve cross-sectional imaging (spiral computed tomography and/or fluorodeoxyglucose positron emission tomography) starting 3 months after surgery and thereafter twice annually. Long-term follow-up for GIST is necessary because there is an unpredictable metastatic pattern.,,,,,,
| » References|| |
Hirota S, Isozaki K, Moriyama Y, Hashimoto K, Nishida T, Ishiguro S, et al
. Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 1998;279:577-80.
Sarlomo-Rikala M, Kovatich AJ, Barusevicius A, Miettinen M. CD117: A sensitive marker for gastrointestinal stromal tumors that is more specific than CD34. Mod Pathol 1998;11:728-34.
Joensuu H, Roberts PJ, Sarlomo-Rikala M, Andersson LC, Tervahartiala P, Tuveson D, et al
. Effect of the tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor. N Engl J Med 2001;344:1052-6.
Demetri GD, von Mehren M, Blanke CD, Van den Abbeele AD, Eisenberg B, Roberts PJ, et al
. Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 2002;347:472-80.
Fletcher CD, Berman JJ, Corless C, Gorstein F, Lasota J, Longley BJ, et al
. Diagnosis of gastrointestinal stromal tumors: A consensus approach. Hum Pathol 2002;33:459-65.
Miettinen M, Lasota J. Gastrointestinal stromal tumors: Pathology and prognosis at different sites. Semin Diagn Pathol 2006;23:70-83.
Rudolph P, Gloeckner K, Parwaresch R, Harms D, Schmidt D. Immunophenotype, proliferation, DNA ploidy, and biological behavior of gastrointestinal stromal tumors: A multivariate clinicopathologic study. Hum Pathol 1998;29:791-800.
Iesalnieks I, Rümmele P, Dietmaier W, Jantsch T, Zülke C, Schlitt HJ, et al
. Factors associated with disease progression in patients with gastrointestinal stromal tumors in the pre-imatinib era. Am J Clin Pathol 2005;124:740-8.
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.
Singer S, Rubin BP, Lux ML, Chen CJ, Demetri GD, Fletcher CD, et al
. Prognostic value of KIT mutation type, mitotic activity, and histologic subtype in gastrointestinal stromal tumors. J Clin Oncol 2002;20:3898-905.
Emory TS, Sobin LH, Lukes L, Lee DH, O'Leary TJ. Prognosis of gastrointestinal smooth-muscle (stromal) tumors: Dependence on anatomic site. Am J Surg Pathol 1999;23:82-7.
Ahmed I, Welch NT, Parsons SL. Gastrointestinal stromal tumours (GIST)-17 years experience from Mid Trent Region (United Kingdom). Eur J Surg Oncol 2008;34:445-9.
Miettinen M, Monihan JM, Sarlomo-Rikala M, Kovatich AJ, Carr NJ, Emory TS, et al
. Gastrointestinal stromal tumors/smooth muscle tumors (GISTs) primary in the omentum and mesentery: Clinicopathologic and immunohistochemical study of 26 cases. Am J Surg Pathol 1999;23:1109-18.
Reith JD, Goldblum JR, Lyles RH, Weiss SW. Extragastrointestinal (soft tissue) stromal tumors: An analysis of 48 cases with emphasis on histologic predictors of outcome. Mod Pathol 2000;13:577-85.
Nilsson B, Bümming P, Meis-Kindblom JM, Odén A, Dortok A, Gustavsson B, et al
. Gastrointestinal stromal tumors: The incidence, prevalence, clinical course, and prognostication in the preimatinib mesylate era: A population-based study in western Sweden. Cancer 2005;103:821-9.
Steigen SE, Eide TJ. Trends in incidence and survival of mesenchymal neoplasm of the digestive tract within a defined population of northern Norway. APMIS 2006;114:192-200.
Tworek JA, Appelman HD, Singleton TP, Greenson JK. Stromal tumors of the jejunum and ileum. Mod Pathol 1997;10:200-9.
Miettinen M, Sobin LH, Lasota J. Gastrointestinal stromal tumors of the stomach: A clinicopathologic, immunohistochemical, and molecular genetic study of 1765 cases with long-term follow-up. Am J Surg Pathol 2005;29:52-68.
Trupiano JK, Stewart RE, Misick C, Appelman HD, Goldblum JR. Gastric stromal tumors: A clinicopathologic study of 77 cases with correlation of features with nonaggressive and aggressive clinical behaviors. Am J Surg Pathol 2002;26:705-14.
Brainard JA, Goldblum JR. Stromal tumors of the jejunum and ileum: A clinicopathologic study of 39 cases. Am J Surg Pathol 1997;21:407-16.
Rutkowski P, Nowecki ZI, Michej W, Debiec-Rychter M, Woźniak A, Limon J, et al
. Risk criteria and prognostic factors for predicting recurrences after resection of primary gastrointestinal stromal tumor. Ann Surg Oncol 2007;14:2018-27.
Wong NA, Young R, Malcomson RD, Nayar AG, Jamieson LA, Save VE, et al
. Prognostic indicators for gastrointestinal stromal tumours: A clinicopathological and immunohistochemical study of 108 resected cases of the stomach. Histopathology 2003;43:118-26.
Lasota J, Dansonka-Mieszkowska A, Sobin LH, Miettinen M. A great majority of GISTs with PDGFRA mutations represent gastric tumors of low or no malignant potential. Lab Invest 2004;84:874-83.
Lasota J, Stachura J, Miettinen M. GISTs with PDGFRA exon 14 mutations represent subset of clinically favorable gastric tumors with epithelioid morphology. Lab Invest 2006;86:94-100.
Miettinen M, Makhlouf H, Sobin LH, Lasota J. Gastrointestinal stromal tumors of the jejunum and ileum: A clinicopathologic, immunohistochemical, and molecular genetic study of 906 cases before imatinib with long-term follow-up. Am J Surg Pathol 2006;30:477-89.
Nada R, Vaiphei K, Singh R, Wig JD, Gupta NM, Banerjee AK. Gastrointestinal stromal tumors the assessment of malignant potential. Indian J Pathol Microbiol 2001;44:393-7.
Nakamura N, Yamamoto H, Yao T, Oda Y, Nishiyama K, Imamura M, et al
. Prognostic significance of expressions of cell-cycle regulatory proteins in gastrointestinal stromal tumor and the relevance of the risk grade. Hum Pathol 2005;36:828-37.
Yasui H, Konishi E, Urasaki K, Yasukawa S, Yanagisawa A. Intratumoral heterogeneity of MIB-1 labelling index in gastric gastrointestinal stromal tumor (GIST). Gastric Cancer 2006;9:223-8.
Kim KM, Kang DW, Moon WS, Park JB, Park CK, Sohn JH, et al
. Gastrointestinal stromal tumors in Koreans: It's incidence and the clinical, pathologic and immunohistochemical findings. J Korean Med Sci 2005;20:977-84.
[Figure 1], [Figure 2], [Figure 3], [Figure 4]