Surgical resection for gastrointestinal stromal tumors (GIST): experience on 25 patients
© Boni et al. 2005
Received: 27 September 2005
Accepted: 30 December 2005
Published: 30 December 2005
Gastrointestinal stromal tumors (GIST) are infrequent and diagnosis and prognosis could be troublesome. We present short and long term results of surgical resection for GIST at the Department of Surgery, University of Insubria, during a period of 17 years.
Materials and methods
All patients' data, tumor characteristics, surgical procedure and survival data were analyzed retrospectively. Tumors were divided in risk classes using the classification proposed by Fletcher, based on tumor size and number of mitosis.
Between 1987 and 2004, 25 patients underwent surgical resection for GIST. Stomach was the most common site of localization. Complete resection was achieved in 88% cases, while in 12% radical resection was not possible. The mean tumor size was 9.2 cm (1.2 - 30 cm): <5 cm diameter in 14/25 cases (56%), 5–10 cm in 5/25 (20%) and >10 cm in 6/25 (24%). Mitotic count was <10/50 HPF in 68% (17/25) and >10/50 in 32% (8/25). Using Fletcher's classification, tumors were divided in very low (11/25, 44%), low (4/25, 16%), intermediate (6/25, 24%) and high-risk (4/25, 16%) groups. The 5-year overall survival was 65% and 34% respectively with a statistically significant difference between tumors <5 cm and >10 cm in diameter and between complete and incomplete resection. High-risk tumors had a significantly shorter survival than low or very low risk.
Our experience confirms that GIST's are uncommon and aggressive cancers. The prognosis is strictly related to tumor size and number of mitosis. Although significant advances on new chemotherapeutic regimes have been made, to date, only radical surgery offers the chance of long-term survival.
Gastrointestinal stromal tumors (GIST) are extremely unusual neoplasm accounting for less than 1% of all gastrointestinal tumors, arising from the Cajal's interstitial cells located in mesodermal tissue. They are defined as primary mesenchymal tumors, typically staining positive for the expression of c-KIT protein . Diagnosis is complex and always requires immunohistochemical staining, since it is based on specific ultrastructural characteristics and positivity for specific immunophenotype markers [1, 2].
Overall survival after surgical resection and clinical behavior of GIST are depending to tumor size and mitotic count, regardless their benign or malignant microscopic features . Fletcher et al  divided GIST in different classes of risk, based on the analysis of prognostic factors [4, 5] classifying these tumors in different "risk classes".
Despite recent advances in chemotherapeutic regimens, such as introduction of targeted therapy with inhibitors of tyrosine kinase receptors , surgical resection is still considered the treatment of choice for GIST .
We present short and long-term results of surgical resection of patients suffering from different types of GIST over a period of 17 years.
Materials and methods
Between 1987 and 2004, 25 patients (15 male and 10 female, mean age of 69 ± 8 years, range 34–80) underwent surgical resection for different type of GIST. The tumors were located in the stomach, small bowel, large bowel and peritoneum in 10, 10, 2 and 3 patients respectively.
All patients' data, tumor characteristics, surgical procedure, intra and postoperative complications, as well as follow-up and survival data, was entered in a specifically designated computer data base and was analyzed retrospectively.
Histopathological examination of surgical specimens was carried out using standard hematoxylin and eosin staining as well as specific immunohistochemical techniques allowing the identification of tumor's grade (low, moderate, and high), size, cellular pattern, stromal background, stage and number of mitosis at high-power field (HPF).
Using the standard histological classifications , 9/25 (36%) tumors were considered malignant, 2/25 (8%) of uncertain behavior (border line) and 16/25 (64%) benign.
The mean tumor size was 9.4 cm (range 1.2 - 30 cm): <5 cm diameter in 14/25 (56%) cases, 5–10 cm in 5/25 (20%) and >10 cm in the remaining 6/25 cases (24%).
Mitotic count was low (<10/50 High-Power Field) in 68% (17/25) of the tumors and was high (>10/50 High-Power Field) in the remaining cases (32% 8/25).
Proposed approach for defining risk of aggressive behavior in GIST, by Fletcher 
Tumor Size (cm)
Very low risk
< 2 cm
< 5/50 HPF
< 5/50 HPF
< 5 cm
< 5/50 HPF
> 5 cm
> 5/50 HPF
> 10 cm
any mitotic rate
> 10/50 HPF
Five out of 25 patients (20%) were affected by significant coexisting malignancy: mesocolic located tumor associated with lung cancer (one patient); adenocarcinoma of the stomach (three patients) and, finally, cancer of the left colon was associated to pelvic GIST (one patient). These patients, even if both tumors were resected, were excluded from the survival analysis.
Clinical follow-up recorded late postoperative complications, recurrence (local and distant) and actuarial survival based on Kaplan-Meier method.
Surgical procedures for resection of GIST
Local resection = 6 (60%)
Partial gastrectomy = 3 (30%)
Local resection + excision of diaphragmatic tumor deposits = 1 (10%)
Small bowel (10)
Local resection = 4 (40%)
Small bowel resection = 6 (60%)
Local resection = 1 (33.3%)
Bulky excision = 2 (66.7%)
Local resection = 1 (50%)
Left colon resection + pelvic metastasectomy = 1 (50%)
In 3/25 (12%) cases, radical resection was not possible due to diffuse disease (2 cases) or local infiltration (1 cases) and the final histopathology described the specimens as malignant, high-risk GIST (Table 1). There was no intra or postoperative mortality. In one case (5.2%) accidental intraoperative spleen damage occurred that was treated by splenorraphy.
Postoperative mortality rate was 32% (8/25). Four cases died of wound infection, two of lung atelectasis, one of pleural effusion and one of central venous catheter infection. The mean follow-up period was 54 ± 10 months (range 22–120).
Gastrointestinal stromal tumors are rare neoplasm that may arise virtually in any part of the gastrointestinal tract. It has been recently proposed that GISTs may develop from the interstitial cells of Cajal , since the same immunohistochemical markers can identify GIST. Presence of CD34 antigen and the c-KIT proto-oncogene (CD117) represent the histological features of these tumors [1, 3, 8].
Other studies [1–3, 8] also demonstrated the diagnostic role of c-KIT expression, now considered a highly specific marker for GISTs. c-KIT expression can also be used for medical therapy , since new drugs, (STI-571 imatinib), is characterized by a selective action on tyrosine kinase receptors, and it has been recently used with good results [6, 12, 13].
C-KIT always immunostains positively in GISTs, with very rare exceptions, due to artifacts, sampling errors, lack of kit caused by clonal evolution or mutations . Furthermore c-KIT expression is extremely helpful to validate diagnosis of GIST in case of extra-gastrointestinal localization (omentum, mesentery, retroperitoneum or elsewhere in the abdominal cavity) .
Brennan et al, in 200 patients found no correlation between tumor's site of origin and survival. Opposite results have been reported by Lillemoe et al , in 133 resected GIST, where survival was, indeed in relationship with tumors' site.
There are difficulties to classify benign or malignant GIST using the standard criteria commonly used for other tumors and most authors agree that tumors size and the number of mitoses at HPF are the most important factors related to prognosis [2, 3, 5]. These findings allowed Fletcher et al , to propose a "risk of aggressive behavior" classification of GIST considering only size and HPF mitotic count (Table 1).
Surgical resection is still "the gold standard treatment" for GISTs, allowing to reach a cumulative 3–5 years survival of almost 50% and 35% respectively [1, 3, 4]; these findings are confirmed by our series (Figure 1).
Our study confirms that size of the tumors as well as mitotic count are highly related to the prognosis: patients suffering from GISTs less than 5 cm in diameter have a significantly longer survival than patients with bigger tumors: 126 ± 23 months for tumor <5 cm, 81 ± 16 for 5–10 cm and 43 ± 12 for >10 cm GIST (Chart 2) (p < 0.05).
Using the "risk of aggressive behavior classification" (Table 1)  for our patients, we confirmed that low number of mitosis at HPF are related to prognosis: we found a significantly longer survival in very low and low risk group compared with high risk group (p < 0.05).
In our experience macroscopically complete resection was possible in 21/25 cases (84%) (Table 2); the presence of residual tumor was significantly related to early recurrence and short survival (Figure 3).
The negative effect of macroscopic residual tumor is well known: Ott et al,  and Brennan et al,  found a significantly longer 5 year survival rate when GIST were completely removed (42% versus 9%).
We found no difference in term of survival considering different tumor locations, but the prognostic significance of the site of origins in term of survival is still controversy. While Brennan et al's , finding are in keeping with ours, results from the 164 patients treated at the Johns Hopkins Hospital showed longer survival for esophageal and duodenal located GIST .
Since approximately 50% of the patients will not survive more than 5 years even when complete resection is achieved, several chemo-radiotherapeutic adjuvant regimens have been proposed.
As these tumors mainly arise in gastrointestinal tract and irradiation may produce severe damage on adjacent organs, radiotherapy seems to be indicated only to prevent local recurrence in selected cases of rectal GISTs [7, 11].
A combined European/Australian study [2, 12] has been started, in order to test the effectiveness of a new chemotherapy regimen based on imatinib mesylate STI-571: a tyrosine kinase receptors inhibitor , highly selective on GISTs and able to perform a "target" therapy, based on the c-KIT expression (CD117) of these tumors. Preliminary results of the first phase trials demonstrated up to 80% of partial response in patients suffering from metastatic or recurrent GISTs treated by imatinib.
An adjuvant or palliative therapy by oral administration of imatinib is now considered mandatory for unresectable or metastatic diseases as suggested by a recent and extensive review by Bucher et al , while its role in potentially resectable GITS or as neo-adjuvant regimes has to be demonstrated yet.
Our experience confirms that GISTs are uncommon and aggressive cancers. The prognosis is strictly related to size and number of mitosis at HPF. Complete surgical resection of the tumors still remains the only chance of long term survival for these patients. Imatinib therapy should be used for patients not suitable for surgery due to poor general condition or once a complete resection it is not technically possible. Its role in adjuvant settings after complete excision is still controversial.
- Trupiano J, Stewart R, Misick C, Appleman H, Goldblum J: Gastric stromal tumors, a clinicopathological study of 77 cases with correlation of features with nonaggressive and aggressive clinical behaviour. Am J Surg Path 2002, 26:705–714.View ArticlePubMed
- Fletcher C, Berman J, Corless C, Gorstein F, Lasota J, Longley BJ, Miettinen M, O'Learly TJ, Remotti H, Rubin BP, Shmookler B, Sobin LH, Weiss SW: Diagnosis of gastrointestinal stromal tumors: a consensus approach. Hum Pathol 2002, 33:478–483.View Article
- Miettinen M, El-Rifai W, Sobin LH, Lasota J: Evaluation of malignancy and prognosis of gastrointestinal stromal tumors: a review. Hum Pathol 2002, 33:459–465.View ArticlePubMed
- DeMatteo R, Lewis JJ, Leung D, Mudan S, Woodruff J, Brennan MF: Two hundred gastrointestinal stromal tumors: recurrence patterns and prognostic factors for survival. Ann Surg 2000, 231:51–64.View ArticlePubMed
- Bilimoria M, Holtz DJ, Mirza NQ, Feig BW, Pisters PW, Patel S, Pollock RE, Benjamin RS, Papadopoulos NE, Plager C, Murphy A, Griffin JR, Burgess MA, Hunt KK: Tumor volume as a prognostic factor for sarcomatosis. Cancer 2002, 94:2441–2446.View ArticlePubMed
- Shawer LK, Slamon D, Ullrich A: Smart drugs: tyrosine kinase inhibitors in cancer therapy. Cancer Cell 2002, 1:117–123.View Article
- Pierie JP, Choudry U, Muzikansky A, Yeap BY, Souba WW, Ott MJ: The effect of surgery and grade on outcome of gastrointestinal stromal tumors. Arch Surg 2001, 136:383–389.View ArticlePubMed
- Di Matteo G, Pescarmona E, Peparini M, DiMatteo FM, Zeri KP, Mascagni D, Mele R, Maturo A, Redler A, DeAntoni E: Histopathological features and clinical course of the gastrointestinal stromal tumors. Hepatogastroenterol 2002, 49:1013–1016.
- Skandalakis JE, Gray SW: Smooth muscle tumors of the alimentary tract. In: Progress in clinical cancer. Grune & Stratton, New York 1965, 692–708.
- Lillemoe KD, Efron DT: Gastrointestinal stromal tumors. In Current surgical therapy. (Edited by: Cameron JL, Gery L). Mosby Inc, USA 2001, 112–117.
- Shiu MH, Farr GH, Hagdu SI: Myosarcomas of the stomach: natural history, prognostic factors and management. Cancer 1982, 49:177–187.View ArticlePubMed
- van Oosterom AT, Judson I, Verweij J, Stroobants S, Di Paola ED, Dimitrijevic S, Martens M, Webb A, Sciot R, Van Glabbeke M, Silberman S, Nielsen OS: Safety and efficacy of imatinib (STI-571) in metastatic gastrointestinal stromal tumors: A phase I study. Lancet 2001, 358:1421–1423.View ArticlePubMed
- Joensuu H, Roberts PJ, Salomo-Rikala M, Andersson LC, Tervahartiala P, Tuveson D, Silbeman SL, Capdeville R, Dimitrijevic S, Druker B, Demetri GD: Effect of the tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor. N Engl J Med 2001, 344:1052–1056.View ArticlePubMed
- Bucher P, Villiger P, Egger JF, Buhler LH, Morel P: Management of gastrointestinal stromal tumours: from diagnosis to treatment. Swiss Med Wkly 2004, 134:145–153.PubMed
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.