B. Case 2

49 year old male with history of chronic anemia (Hb = 8.1 g/dL)and recurrent diverticulitis, presented with generalized abdominal pain. Previous episode of diverticulitis was about a year prior to current presentation. CT scan was obtained on initial work up and only mild inflammatory changes were noted around the sigmoid. However a mass was noted inferior to the head of the pancreas behind the colonic mesentery. Endoscopic Retrograde Cholangiopancreatography (ERCP) and Upper GI endoscopy was unremarkable. Small bowel follow through series revealed no luminal compression, although mucosal irregularity was noted in the distal duodenum. Positron emission tomographic scan was used to exclude any extra-abdominal disease.

Since the mass was well defined and not present on the previous years tomographic imaging, surgical resection was contemplated. The mass measured 3 cm in largest dimension. Laparotomy revealed the mass to be arising form the third part of the duodenum and adherent to the pancreas. The ligament of Trietz, proximal jejunum, and fourth part of the duodenum was completely mobilized to gain adequate exposure. Complete resection en bloc could be accomplished only by Whipple’s procedure. There was no residual disease elsewhere in the abdomen. The patient was discharged form the hospital on post operative day 12. At 6 month follow up no further therapy was recommended. He has remained disease free 6 years post surgical resection.

i. Pathology

On gross inspection, the tumor appeared to be arising from the superior aspect of the third part of the duodenum extending behind the head of the pancreas. Final tumor dimension was 3 by 3 by 4 cm and was well encapsulated. Entire capsule was intact, although adherent to the pancreas. Cross section of the tumor revealed areas of patchy necrosis. The tumor had extended up to the mucosa without evidence of any breach in continuity. The mitotic figures were about < 1 per 50 HPF. Histology was consistent with leiomyosarcoma with CD34 staining pattern consistent with GIST. All resected margins were negative.

C. Case 3

52 year gentle man presented with a tumor early satiety, worsening regurgitation ad water brash for about 2 months. Medical history was significant for non insulin dependent diabetes. He had been treated with PPI for about a year for reflux disease. During work up, the hemoglobin was measure to be 7.4 g/dL with a hematocrit of 22%. An EGD was performed which revealed complete narrowing of the first part of the duodenum and the scope could not be passed beyond the pylorus. The pylorus itself appeared to be normal. A CT san revealed a mass about 4 cm in diameter arising from the first part of the duodenum extending into the lesser curvature of the stomach.

Staging work-up was negative with no evidence of metastatic disease.

The patient underwent an elective celiotomy. The tumor was pedunculated and arising form the serosal surface of the first part of the duodenum. Resection was successfully accomplished by performing an antrectomy and resection of the first part of the duodenum. The duodenal stump was closed and continuity was re-established via a Billroth II reconstruction.

The patient developed ileus which resolved spontaneously with naso-gastric decompression. Oral intake was commenced on post operative day 10. He was discharged form the hospital on Day 14. Follow up at 15 days and 2 months after surgery was unremarkable.

i. Pathology

Gross dimensions of the tumor were 4 by 5 by 3 cm, mostly arising from the serosal aspect of the first part of the duodenum. The entire luminal aspect of the first part of the duodenum was compress with 0.5cm ulceration in the duodenal bulb. Staining pattern is indicated in Table 1. Both spindle cell and epithelial cell components were observed. Mitotic figures were 7 per 50 HPF. The patient underwent concurrent treatment with Imatinib for a period of six months. There was no evidence of recurrent disease at 8, 12 and 24 months after follow up and he been doing well since.

III. Discussion

Gastrointestinal stromal tumors have been recognized more frequently in the last decade as a clinicopathologic distinct group of tumors with a varying spectrum but characterized by distinct immmunohistochemical patters. GIST are characterized by a vast heterogeneity in cellular origin, differentiation, clinical behavior, anatomic distribution and prognosis - most of which had been poorly understood until the past 5 years (Lee JS et al, 1995; Graadt van Roggen et al, 2001). Initially confused with smooth muscle tumors (Lavin et al, 1972; Miettinen, 1988), they were subsequently individualized based on ultrastructural and immuno-histopathologic patterns.

Our aim is to review in detail the management of GIST that arises from the duodenum or its vicinity. We define any tumor in contact with the duodenum, based on any imaging, as a tumor of the pancreatoduodenal complex. GISTs have been referred to by various acronyms, including GANT (gastric autonomic nerve tumors), STUMP (smooth muscle tumors of uncertain malignant potential) and GIPACT (gastrointestinal pacemaker cell tumor). The many acronyms only serve to highlight the controversy associated with the cellular origin of these tumors (Dierkes-Globisch et al, 2001).

A. Defining GIST

GISTs are best defined as a group of neoplasms of non-epithelial origin. Located along the entire length of the gastrointestinal tract, they arise either from epithelioid or spindle cells whose embryologic origins are shared with the Interstitial Cells of Cajal (ICC). They manifest along a spectrum ranging from those involving predominantly neural components to those with predominantly myoid components (Neuhaus et al, 2005). The term GIST was introduced in 1983 (Mazur and Clark, 1983) and for many years remained a vaguely defined type of tumor. In 1984, Herrera et al. described a subgroup of gastrointestinal stromal tumors termed GANT (gastric autonomic nerve tumors) or plexosarcomas (Herrera et al, 1984) based on ultra structural properties. Consequent to their similarity to autonomic nerve structures, they were presumed to originate from the intramural autonomic plexus.

Kindblom and colleagues subsequently demonstrated in 1998 that the ICC of the gastrointestinal tract form a complex cellular neural network, which is proposed to play a vital role in intestinal motility. These cells exhibit both myoid and neural features, and were presumed to be candidates for the histogensis of the tumor. However, in detailed reviews that followed, GANT’s were reclassified as GIST’s. The discovery of the CD 117 (encoded by the KIT gene) gain of function mutation by Hirota and colleagues in 1998 was an important landmark responsible for the reclassification of these tumors into a distinct group. The previous GANT type of gastrointestinal stromal tumor were presumed to be a more aggressive variant of GIST (Tornoczky et al, 1999). Although immunohistochemistry may be helpful for classifying GIST, electron microscopy is considered the gold standard for making a diagnosis of GANT, which is now classified as a subdivision of GIST and managed similarly (Segal et al, 1994).

GIST’s lack the distinct features of classic leiomyomas, schwannomas and their malignant counterparts (Kerr et al, 1999). Although the terms leiomyoblastoma and plexosarcoma (Appelman and Helwig, 1977; Dalaker and Harket, 1980) were used depending on the predominant cellular pattern, the inclusive term GIST was introduced to collectively refer all these tumors (Saul et al, 1987). The discovery of the CD 34 stem cell antigen expression in these tumors was an important landmark and used as a major diagnostic criterion for GIST (van de Rijn et al, 1994; Miettinen et al, 1995). The observation of a more consistent expression of the KIT gene in a series of GIST (Ernst et al, 1998) led to a completely different approach to these tumors. Gain of function mutations in the juxtamembranous domain of the KIT gene results in a ligand independent activation of the CD 117 receptor (Hirota et al, 1995). Studies have correlated KIT gene mutations with the malignant potential of CD 117 expressing tumors (Lasota et al, 1999; Taniguchi et al, 1999).

B. Interstitial cells of Cajal and CD 117 mutation

Torihashi and colleagues demonstrated in 1999 that blockade of KIT gene signaling induces transdifferentiation of interstitial cells of Cajal to a smooth muscle phenotype, indicating the plasticity between ICC and smooth muscle cells and the common origin based on in vitro studies. The same mechanism is proposed to be responsible for the etiopathogenesis of GIST (Pauwels et al, 2005).

Huizinga and colleagues showed in 1995 that the KIT gene variants were responsible for both ICC and intestinal pacemaker activity. CD 34 expression is noted in many cell types and hence CD 34 expression in GIST does not define lineage but merely suggests the primitive nature of the mesenchymal cells (Miettinen et al, 1995). Although their functional aspect remains controversial, some authors still believe the ICC to be an integral part of gut motor physiology, controlling peristalsis and muscular activity by acting as pacemaker cells and possibly as neurotransmitter mediators (Huizinga et al, 1997; Sircar et al, 1999). ICC and smooth muscle cells have been shown to originate from common precursors (Kluppel et al, 1998). Other features identifying GIST include ultra structural identification of complex interdigitating cell processes with rudimentary cell junctions (Rumessen, 1996).

The ICC do not have unique identifying characteristics: identification is dependent on ultra structural features and their spatial relationship to surrounding cells, which is unfortunately distorted or lost in neoplatic cells (Sircar et al, 1999). Although they have close contact with nerve bundles and smooth muscles, their contractile apparatus is less well organized than in smooth muscles (Faussone-Pellegrini et al, 1990). The ligand for the membrane-bound tyrosine kinase receptor CD 117 encoded by the KIT gene happens to be stem cell factor (SCF). The interaction between SCF-CD 117 is important elsewhere for the maturation of germ cells, bone marrow stem cells, melanocytes and mast cells, and is not restricted to the ICC family (Huizinga et al, 1995). In normal gut, the ICC are the only cells which demonstrate positivity for CD 117, CD 34 and vimentin. ICC are also MSA negative, desmin negative, PGP negative, S100 negative and tryptase negative (Sircar et al, 1999). This is essential in differentiating GIST from other confounding tumor types. The close differential diagnosis would be ganglia and mast cells which are vimentin positive, CD 117 positive and yet CD 34 negative. Ganglion cells are NSE and S-100 positive. Schwannomas are PGP positive, S-100 positive, vimentin positive and negative for CD-117, CD 34 and MSA. Leiomyomas on the other hand are MSA, desmin and vimentin positive and CD- 117/CD 34/PGP negative (Rumessen et al, 1992).

Gain of function mutation in KIT gene results in activation of the CD 117 without the necessity of ligand interaction. Clone analysis has shown the CD 117 receptor to be only immunoreactive on half of the cells (Hirota et al, 1998). This indicates that only one allele is mutated and responsible for the tumor phenotype, while the normal allele, which is not mutated in the remaining clones, is responsible for the staining characteristics. The mutant allele responsible for the neoplatic phenotype loses immunoreactivity to the CD 117antibodies. Loss of the second allele may result in a more aggressive tumor and increased metastasizing potential. Hence, a tumor arising out of a KIT gene mutation may by virtue of mutation lose reactivity to CD -117and yet result in a gain of function mutation, thereby presenting as a clinically aggressive CD 117 negative stromal tumor (Hirota et al, 1998). A similar analogy to CD 34 reactivity has been proposed. Loss of CD-117 and CD 34 immunoreactivity within the same tumor is responsible for the variation in stromal characteristics of these immunophenotypic markers. Hence GIST’s, although most commonly CD 34 positive and CD 117 positive, could still present as a variant form, with either one or both markers staining negative. An aggressive form of GIST could thus be missed on immunohistochemical staining if all the alleles of CD 117 or CD 34 are mutated although this happens to be a very rare variant. CD 34 negativity without loss of reactivity to CD 117 happens to be the more common aggressive variant (Hirota et al, 2006). Clinically this is important when dealing with very aggressive types of GIST which lack consistent staining patterns, making diagnosis and treatment difficult.

Platelet-derived growth factor receptor α (PDGFRα) has been shown to be mutated in certain types of GIST without the evidence of a KIT gene mutation (Bernet et al, 2003). Tumor cells could be non-reactive to CD-117 antibodies due either to non-expression or mutation of KIT gene resulting in expression of different epitopes of CD 117 which are not recognized by the CD-117antibodies. The latter case could be a gain of function mutation resulting in over expression of KIT gene in the setting of non-reactive immunohistochemistry (Miettinen and Lasota, 2006). Mutations in the juxtamembranous domain of CD-117 (exon 11 and 9) are the most common in GIST’s of all sites (Buchdunger et al, 2000). No specific exon mutation has been defined for duodenal GIST due to the rarity of cases indicating their potentially different behavior. KIT gene mutations leading to ligand-independent activation of the receptor, resulting in uncontrolled intracellular phosphorylation, are seen in 50 % of GIST’s (Hirota et al, 1998). Specific treatment of GIST relies on the utilization of tyrosine kinase inhibitors to block the over activity of the mutated CD 117 receptor and is the basis for current drugs utilized in adjuvant therapy (Lee et al, 2001).

C. Sub classification of gastrointestinal mesenchymal tumors

In a series of 244 cases (Rudolph et al, 2002), antigen expression varied from tumor to tumor except for vimentin, which was consistently positive. CD 117 positive tumors were classified as GIST regardless of the co-expression of the other antigens. CD 117 negative tumors with strong expression of myogenic markers, and with no significant neural or CD 34 expressions, were considered GI leiomyogenic tumors. CD117 negative tumors with strong S-100 or GFAP or PGP positivity in the absence of myogenic tumor markers were considered GIGT (gastrointestinal glial/schwannian tumors). Concurrent expression of synaptophysin in CD 117 negative tumors was characterized as GINT. GIFT’s were a special category of tumors with null phenotype lacking any antigen expression except for vimentin. Strong CD 34 positive tumors that were CD 117 negative and did not fit any of the above criteria were classified as GINST (gastrointestinal negative stromal tumors) (Table 2). However, if the previous hypothesis that mutation of both KIT gene alleles results in absence of staining but in a more aggressive form of tumor, then GINSTs should indeed be a more aggressive form of GIST. The data presented by Rudolph and colleagues in 2002 and DeMatteo and colleagues in 2000 is consistent with that hypothesis.

The percentage of tumor phenotypes varied according to the anatomic distribution of the tumor with GIST accounting for 100% of esophageal cases, 80 % of jejunal tumors and 54% of gastric tumors. No specific data is available for duodenal tumors. On an overall basis, 9% of all GINST’s occurred in the duodenum. GIFT’s were almost exclusively noted in the duodenum. Overall, 9.8 % occurred in the duodenum, with GIST’s and GIFT’s being the most common forms (Rudolph et al, 2002).

Tumors categorized as GIFT’s, if assumed to be dual mutant forms of CD 34 and CD 117 would technically have a worse prognosis. The cumulative survival rate for GIST’s has been reported as around 0.6, GINST’s as 0.4 and GIFT’s as 0.2. Hence, CD-117 negative tumors, although classified as a separate group, could in fact be more aggressive forms of GIST, involving mutations of both the KIT gene alleles. Genetic analysis for KIT mutations would be more helpful than immunohistochemistry in characterizing the nature of GIST’s. Of note is the fact that CD 117 negative stromal tumors exhibit more epithelioid features (Tortella et al, 1987).

D. Malignant potential and mitotic figures

Traditionally a mitotic index of more than 4 mitotic figures per 10 high power fields (HPF) was considered malignant. However, recent reports have proposed a revised mitotic index of more than 10 mitotic figures / 50 HPF to be more appropriate (Meesters et al, 1998). In a classification proposed by Ranchod and Kempson in 1977 for grading GIST, a tumor more than 4 cm in size was classified as a gastrointestinal sarcoma if it was hypercellular and associated with a high mitotic index. These tumors were considered low grade if the mitotic index was less than 10 per 50 HPF, and high grade if more than 10 per 50 HPF.

Recent work by Fletcher and colleagues in 2002 has reclassified the malignant potential of GIST tumors based on size and is currently held as the standard. Cut off values for tumor sizes are 5cm and 10cm, while mitotic figures are 5 and 10 per 50 HPF. Tumors which have a mitotic index of more than 10 /50 HPF or a size of more than 10 cm or those which are intermediate in size (5 to 10 cm) with a mitotic index between 5 and 10 /50 HPF were all classified as high risk tumors. Tumors with a size of less than 5cm and a mitotic index of less than 5 per 50 HPF were considered low risk. A subset of this group with size less than 2 cm was classified as very low risk. All other combinations were classified as intermediate risk tumors. This classification, although useful in directing the indication for adjuvant therapy, does not help in identifying metastatic potential. Even tumors of low risk can potentially metastasize, thereby highlighting the point that all GIST's should be treated with caution (Schubert et al, 2006). The other problem associated with this risk stratification is that the mitotic grade will be known only after the final specimen is resected, and hence will not be helpful in guiding the extent of surgical resection preoperatively. Surgical resection has to be decided primarily on size and feasibility of resection based on the anatomic location of the tumor. A fine needle aspiration biopsy (FNAB) could be very helpful in confirming the diagnosis, but may be very difficult to use in stratifying low or intermediate grade tumors. The presence of mitotic figures in a FNAB is more likely consistent with a high risk tumor (Elliott, 2006).

E. Anatomic distribution

GIST accounts for most mesenchymal tumors within the gastrointestinal tract (Evans et al, 1985), highlighting the non-epithelial nature of the tumor. Tumors of true smooth muscle, neural (schwannoma), fibroblastic and vascular origin are thus excluded. Differentiating from pure smooth muscle cell tumors is an important aspect in the diagnosis of GIST. Smooth muscle tumors show variably spindled cells with cigar shaped nuclei and bipolar perinucler location of the cytoplasmic glycogen. Smooth Muscle Actin (SMA) and Muscle Specific Actin (MSA) are generally positive in smooth muscle tumors but patchy or absent in GIST (Bagnolo et al, 1998).

GIST’s can arise in any part of the GI tract, but more commonly occur in the stomach (65 to 70%) and small bowel (30 to 45%), and less frequently in the esophagus, colon and rectum where true myogenic tumors predominate (Antonioli et al, 1989). Pure leiomyomas are less common than GIST’s in the duodenum, and the esophagus is the only location in the GI tract where leiomyomas predominate compared to GIST (Emory et al, 1999). The majority of GIST’s are located in the stomach and small intestine with only 4% located in the duodenum (Meesters et al, 1998). A recent review paper indicated a higher percentage in the duodenum but only in a single institution (Winfield et al 2006). Also, some of the retroperitoneal GIST tumors were found in a juxtaduodenal location, further indicating that these, in fact, might be arising from the duodenum itself (Lee et a; 2001).

F. Clinical presentation

Overall, the peak incidence of GIST is in the fifth and sixth decades, being infrequent before the age of 40 (Hinz et al, 2006). Based on a review of the cases reported in the literature (Sakamoto et al, 2003; Hughes et al, 2004; Winfield et al, 2006) GIST’s of the pancreatoduodenal complex typically present with bleeding. Most small GIST’s are asymptomatic and go unnoticed until an esophagogastroduodenoscopy is performed for some other reason. Symptomatology is dependent on the size and the location of the tumor (Hompes et al, 2004). The most common complaint for patients with GIST is usually vague abdominal pain or discomfort (De Marco et al, 2005). This is unlike the presentation noted in GIST of the duodenum, which typically manifests with bleeding and associated microcytic anemia (Carvajal et al, 2006).

All three cases encountered in our institution presented initially with anemia. We attributed the bleeding to pressure erosion of the duodenum. As opposed to other areas of the GI tract, GIST’s arising from the duodenum do not have surrounding free space to grow, making the pliable duodenum the least resistant pathway for tumor expansion. Other presentations include anemia, anorexia, weight loss, nausea, fatigue, and acute intraperitoneal bleeding or perforation (Goh et al, 2005). The tumor may or may not arise from the duodenum per se, but rather from the retroperitoneum or any adjacent structure.

As with most primary mesenchymal tumors, GIST’s of the pancreatoduodenal complex tend to be typically of submucosal origin, indicating that they have to be associated with the duodenum in most circumstances with extension possible into any of the surrounding structures. GIST’s tend not to infiltrate but to push surrounding structures away. That being said, the desmoplastic reaction in large tumors could make it technically difficult to identify the exact tissue plane. Pressure necrosis associated with the tumor pushing onto the surrounding structures could result in obliteration of the capsule due to inflammatory response. Based on classification by Yamada and Ichikawa in 1974, most tumors are Type 1 tumors, i.e. sessile or slightly raised. Rare reports of Type IV - polypoid lesions with long stalks arising from one region of the GI tract and causing a distal obstruction down stream could be found in the literature (Kim et al, 1999).

G. Diagnostic and staging workup

With the initial presentation usually being an upper GI bleed, EGD is the first modality of evaluation. If the EGD reveals an obvious lesion, then a luminal biopsy should be considered. In most circumstances, GISTs are not diagnosed until there is a final pathologic examination of the resected specimen. GIST tumors with erosion into the lumen of the duodenum are easily accessible for tissue biopsy and a preoperative diagnosis can be established in some cases. In most circumstances, the tumors can be however easily approached via an Endoscopic Ultrasound guided Fine-needle aspiration (EUS-FNA). EUS-FNA has advantages in that intraperitoneal seeding of the tumor is avoided and a diagnosis can be established with reasonable certainty (Chatzipantelis et al, 2008). Adequate specimen sufficient enough to perform the appropriate immmunohistochemical stains could be easily obtained. It has been shown that that when combining cytologic and immunocytochemical studies, EUS-FNA is accurate and efficient in the diagnosis of GIST (Fu et al, 2002). However, in those cases where the tumor is not easily accessible for a transgastrointestinal luminal biopsy, a high index of suspicion is necessary. In such circumstances, transperitoneal biopsy is best avoided due to tumor spillage associated with the procedure. Instead a detailed non-invasive staging work up must be considered. Initial diagnostic work up consists of a contrast-enhanced CT scan with reconstruction, if feasible. The relation of the tumor to the surrounding structures has to be delineated.

CT scanning is considered the imaging modality of choice for anatomic evaluation of GIST (Lee et al, 2004). Numerous characteristics are assessed, but the most important features indicating a need for further workup are the sharpness of the tumor margin and its proximity to the luminal aspect of adjacent duodenum. Evaluating the extent of the mass, detecting possible metastasis, and assessing the resectability of disease during staging are easily accomplished utilizing a CT scan. A triphasic scan should be obtained for baseline evaluation (Blay et al, 2005). Nodal evaluation is unnecessary as these tumors very rarely involve lymph nodes. If tumors are juxtaduodenal, then an endoscopic ultrasound would be necessary to evaluate the interface between the tumor and the duodenum. If the tumor is abutting the duodenum, then a transluminal EUS-guided biopsy should be considered without the risk of contaminating the peritoneal cavity.

A sharp interface noted on the CT scan indicates that the tumor may not be infiltrating the surrounding structures, allowing enucleation as a possibility if the tumor is small. Tumors that do not have a sharp delineation are most likely to require a radical approach. 18-Fluodeoxy-glucose Positron emission tomography (18 FDG-PET) may be used to evaluate the extent of the disease and to screen for metastatic disease. However, a PET scan should not be used for initial detection as it is very non-specific: it should be utilized only after previous evaluation with a CT scan. PET has a specific role only in staging a GIST. Since PET and CT scans have different uses, both should be obtained in all cases of duodenal GIST even if the primary tumor is very small. Even tumors with very low risk have a metastatic potential so PET is indicated in these cases also. In circumstances where small GIST’s are missed by CT scan, a PET scan would raise the index of suspicion by highlighting the area of metabolic activity. In addition, a PET scan also delineates the presence of surrounding inflammation due to hemorrhage or myxoid degeneration from that of tumor growth and infiltration (Goerres et al, 2005).

PET and CT scan in combination precisely delineate lesions and allow intricate planning of the surgical procedure. The assessment of operability is a crucial step in the preoperative workup of a GIST, as the prognosis is excellent after complete resection (Emory et al, 2005). PET and CT are important in the current era of imatinib mesylate neoadjuvant treatments, as they help to categorize the disease as responsive, stable or progressive. However, absence of PET activity does not necessarily correlate with absence of tumor cells on pathology. It is also important to analyze the nature and appearance of these lesions after imatinib therapy. Appearance of new lesions may be difficult to identify since liver metastases that are initially occult, could present as new cystic changes due to tumor destruction consequent to adjuvant therapy on a follow up CT scan (Linton et al, 2006). The appearance of a nodule within a mass pattern is an ominous sign and consistent with recurrent disease or new onset resistance. Hence attention to detail and homogeneity of the lesions and not the size alone is crucial in determining response and early identification of resistance to treatment on follow up CT imaging (Shankar et al, 2005).

H. Surgical management

The aim of surgery in any type GIST is to achieve negative microscopic margins. Surgery is the principal modality of treatment in patients with resectable GIST (Pierie et al, 2001; Rossi et al, 2003). Grossly negative margins are not prognostic indicators in GIST. Tumors of the pancreatoduodenal complex pose many technical challenges because of the surrounding anatomy. Unlike tumors of the small intestine or stomach where wide margins can be achieved without a significant technical challenge, tumors in the duodenum do not lend themselves to a simple method of operative resection. The issue to be addressed is the necessity of a pancreaticoduodenectomy for these tumors. The morbidity associated with pancreatoduodenectomy is quite high, so the indications for this procedure in GIST tumors must be specific. There have been reports of even large tumors being treated with wedge excisions especially if the tumor is predominantly extra luminal (Sawaki et al, 2003). In the case described by Sawaki and colleagues 2003, recurrent metastatic disease was noted within two years for a tumor that was resected from the duodenal bulb. In a case series described by Winfield and colleagues 2006, 5 out of 8 cases were treated by pancreaticoduodenectomy, but none of the tumors were located in the first portion of the duodenum.

The question remains: Is pancreaticoduodenectomy the ideal procedure or could less radical approaches be entertained? To answer this we need to better understand the tumor biology. Based on a review of literature, there have been no consistent long term studies in answer to this question. There have been anecdotal cases where the surgical modality has varied from a simple enucleation to a radical pancreaticoduodenectomy. Based on a review by Piere and colleagues in 2001, complete resection is the only modality of cure. Even with complete resection five-year survival was only 42 % and incomplete resection was associated with a 9% survival. In addition, complete resection was achieved in only 59% of cases. The objective of surgery for primary GIST is the resection of tumor with an intact pseudocapsule. This is critical, as damaging the pseudocapsule may allow for bleeding/hemorrhage and/or dissemination. One might favor an argument that enucleation could be utilized as an option as these tumors do not invade the surrounding structures. GIST’s typically are very friable and enucleation is associated with the risk that the pseudocapsule could be violated. Violation of the pseudocapsule could potentially result in spillage of tumor cells and henceforth upstage the disease. With the potential of “cure” only achievable through surgery, complicated by the fact that chemoradiation is not very helpful in GIST, we feel that enucleation should not be attempted at all in GIST. Initial studies by De Matteo and colleagues in 2000 have indicated that although size correlated with poorer prognosis, positive microscopic margins did not. However, subsequent data by Gold and colleagues in 2007, have demonstrated that R0 resection for GIST was associated with a statistically significant improvement in survival compared to either R1 or R2 resection. We strongly favor negative microscopic margins to be accomplished in the very first surgical specimen. The essence here is to avoid traveling though tumor cells at all possibilities and operative dissection should be carried out well away from the vicinity of the tumor.

In certain cases the tumor may be easily lifted off surrounding structures. However, we would like to specify distinctly that only if the tumor is merely sitting on a surrounding structure should an enucleation be considered. This still leaves the pedicle of the tumor to be dealt with. We propose that a 2 cm margin should be obtained at the pedicle. The aim is to achieve negative margins, although it is not clear what the margin should be. Given the small number of cases and the high recurrence rate we feel that a 2 cm margin should be adequate for tumors less than 2 cm in size, or for those with a very low risk potential. If the tumor shows even the slightest evidence of inflammatory adherence to any surrounding structure, one could make a fitting argument not to dissect on the tumor in an attempt to separate it. The friability of the tumor is associated with a high risk of rupture. We prefer a no touch technique by staying away from the tumor even if it means resecting the surrounding structures.

In essence, only very low risk tumors in the free wall of the duodenum can be technically resected by a wide local excision. A tumor in the duodenum of more than 2 cm in size would require more than 6cm of the duodenal wall to be resected (allowing for 2 cm on either side). Reconstruction of such a large defect could be done, but a resection of such magnitude would be associated with unnecessary proximate handling of the tumor. A pancreaticoduodenectomy in such cases would be a better modality as it keeps the dissection well away from the tumor and also provides the highest chance of achieving a negative margin.

There are a few additional factors to be considered before a pancreaticoduodenectomy is performed. The performance status of the patient is important, especially when an operation of such magnitude is considered. A patient with good physiologic reserve, younger age, and no major comorbidity will better tolerate the operation. Such a patient is also more likely to benefit and have the highest chance of surgical cure offered. Tumors confined to the bulb of the duodenum (more than 2 cm proximal to the junction of the first part and second part of the duodenum) could technically be considered for a gastroduodenectomy with a Billroth II reconstruction provided the tumor is on the convex side of the duodenal sweep and well away from the pancreas. Resection of tumors in the fourth part of the duodenum could be accomplished by an intestinal derotation (Valdoni-Strong) procedure (De Nicola et al, 2005). However, if the tumor is located at the junction of the third and fourth part of the duodenum, we strongly suggest that a Whipple’s procedure be performed instead.

As with any major cancer operation, the entire abdominal cavity should be assessed for metastatic disease. The presence of any metastatic disease precludes any major resective procedure, especially a pancreaticoduodenectomy, as the risks undertaken with the operation in these cases are not worth taking since the disease is already advanced. However, palliative procedures may be indicated and biopsy of the suspected metastatic lesions with frozen section confirmation should be done. The role of debulking alone in GIST has not been clearly established. However, as a part of multimodal treatment for advanced disease, debulking has shown to improve survival (Gold et al, 2006).

Raut and colleagues in 2006 reported their data on debulking GIST tumors for cases that were managed with adjuvant imatinib mesylate therapy. Patients were categorized into three groups based on response to tyrosine kinase inhibitor neoadjuvant therapy, namely stable disease, limited progression or generalized progression. After surgery, there was no evidence of disease in 78%, 25%, and 7% of patients with stable disease, limited progression, and generalized progression, respectively. Bulky residual disease remained after surgery in 4%, 16%, and 43% of the patients with stable disease, limited progression, and generalized progression. Twelve-month progression-free survival was 80%, 33%, and 0% for patients with stable disease, limited progression, and generalized progression. Twelve-month overall survival was 95%, 86%, and 0% for patients with stable disease, limited progression, and generalized progression. They concluded that patients with advanced GIST’s exhibiting stable disease or limited progression on kinase inhibitor therapy have prolonged overall survival after debulking procedures. Even though partially successful treatment may render initially inoperable tumors potentially resectable from a technical standpoint, the biologic behavior of the tumor simulates metastatic disease and surgery in these cases is primarily cytoreductive (Neuhaus et al, 2005).

It has been proposed (Rutkowski et al, 2006) that minimizing the tumor burden will decrease the likelihood of imatinib resistant clones from developing within the disease in situ. Surgery has little to offer in the setting of generalized progression. Surgical removal of residual disease after adjuvant therapy may allow complete remission in selected patients responsive to imatinib, theoretically prolonging survival provided imatinib is continued. Simultaneous metastasectomy with resection of the primary in pancreatoduodenal GIST at initial presentation has been reported (Stratopoulos et al, 2006) without neoadjuvant treatment. However, follow-up in this case was not of adequate duration to justify this approach until more data are available on the benefits of neoadjuvant imatinib mesylate.

The paper by Raut and colleagues in 2006 does not specifically address the patient with duodenal GIST or cases where the surgeon discovers advanced disease for the first time at celiotomy. A surprise discovery puts the surgeon in a dilemma as to whether a radical resection should be performed or not. As data are limited, we suggest that if the tumor is of intermediate or low risk, despite the presence of metastatic disease, a debulking operation should be considered. As far as possible a Whipple’s should be avoided: if the patient has a tumor of less than 5 cm in size then one should probably perform a radical resection if the performance status of the patient is excellent. However, this approach remains controversial until more data are available. If the tumor is more than 5 cm then any procedure short of a Whipple’s, which can accomplish a resection of most of the tumor, should be considered. We propose this based on the tumor biology and by extrapolating data from studies that have evaluated options for GIST located in other areas of the GI tract.

Another important factor influencing the prognosis of GIST tumors is the tumor size at presentation. Tumor size also influences the nature of resection to be accomplished, especially if the tumor is located in critical regions such as at the pancreatoduodenal interface. Tumor size at presentation predicts recurrence-free survival: patients with tumors >10 cm have a 5-year survival rate of 27%, while those with tumors <5 cm have 5-year survival of 82% (Raut et al, 2006).

The possibility of discovering advanced disease should be clearly addressed with a pre-operative PET scan. Surgical cures are much more common when incidentally discovered GIST’s are included in the data. These asymptomatic tumors are often small, with benign histologic features, and may be evident only on pathologic examination of a specimen removed for another reason.

I. Laparoscopy

In general, laparoscopy is not indicated once the location of the tumor and the stage has been established. However in doubtful cases where there is a high suspicion of peritoneal disease or liver metastasis, we suggest a laparoscopy and biopsy with intraoperative laparoscopic ultrasound guidance.

J. The role of imatinib mesylate

The phenylaminopyrimidine imatinib mesylate was the product of multiple rounds of selection for a compound capable of inhibiting the Bcr-Abl kinase that is prevalent in CML. The effects of the drug were discovered in 1996 by Druker et al. and have since been utilized in managing many solid tumors (Druker at el, 1996). Imatinib mesylate is a selective tyrosine kinase inhibitor. In addition to Bcr-Abl, imatinib mesylate inhibits KIT and PDGFR. Receptor tyrosine kinases that are not inhibited include EGFR, VEGFR-2, and IGFR (Tuveson et al, 2001; Singer et al, 2002). The mechanism of imatinib mesylate inhibition involves occupying the target kinase’s ATP binding site, denying both the substrate and energy source for tyrosine phosphorylation (Manley et al, 2002).

As the majority of GIST’s are dependent on constitutive KIT gene signaling, experiments performed in GIST tissue culture demonstrated that imatinib mesylate was able to inhibit tissue growth. Until the work by Tuveson and colleagues 2001 GIST was considered a chemo resistant tumor with not many options available for advanced or aggressive disease. The use of imatinib mesylate as a tyrosine kinase inhibitor was primarily restricted to treatment for chronic myeloid leukemia until the work by Tuveson and colleagues where a specific effect was noted on cell lines with oncogenic allele mutations in KIT gene. This correlated with induction of apoptosis and decrease of cellular proliferation. With the first clinical anecdote of imatinib demonstrating significant promise in the treatment of metastatic GIST, GIST began to be included in many randomization trials for various advanced cases and now has been selectively indicated for use in patients with bulky, high grade or advanced disease (Joensuu et al, 2001). Studies have shown up to 90% of patients with metastatic disease stabilizing or responding to imatinib therapy (Demetri et al, 2002). Withdrawal from high dose imatinib therapy has in some cases been associated with a rebound effect leading to tumor flares. This is explained by the high concentration of stem cell factor (SCF) ligand accumulating in the serum (Bono et al, 2004). After therapy is initiated, careful surveillance is necessary to assess the adequacy of drug dosage and tumor response. Currently there are several trials underway by the EORTC and ACOSOG evaluating the efficiency of standard dose imatinib mesylate (400mg a day) with the high dose regimen (800mg per day) (Goldstein et al, 2005). A phase I EORTC study (Demetri et al, 2006), identified the highest feasible dose of imatinib to be 400mg bid and indicated extensive activity against tumor burden. Some studies have shown a small but statistically significant benefit in progression-free survival with the high dose regimen (Verweij at al, 2004).

K. The role of Sunitinib for resistant cases

In patients who demonstrate a disease progression despite imatinib therapy, an alternate drug, sunitinib, which is a selective inhibitor of tyrosine kinase, may be helpful (Demetri et al, 2006). Sunitinib malate is a novel oral multitargeted tyrosine kinase inhibitor with antitumor and antiangiogenic activities. Sunitinib has been recently approved for the treatment of patients with GIST after disease progression or intolerance to imatinib mesylate therapy. Oral sunitinib has been shown to possess a high level of efficacy with the 50 mg daily dose for 4 weeks (Rutkowski et al, 2008). Currently, sunitinib malate is approved only as a second-line drug. Unlike imatinib mesylate, sunitinib is a multitargeted agent inhibiting tyrosine kinase, PDGFR A & B, vascular endothelial growth factor receptors and colony stimulating factor 1 receptor. Resistance to imatinib has been proposed based on mutation of KIT mediated tyrosine kinase receptor ligands and PDGFR α receptors. Sunitinib's efficacy in these circumstances has been attributed to the multitargeted nature of the drug (Le Tourneau et al, 2007). Although this drug is still being evaluated, initial studies have shown a fourfold increase in median time to tumor progression compared to placebo (Joensuu, 2006).

L. Assessment of response to therapy: RECIST versus Choi criteria

CT and PET are both necessary for the monitoring of tumor response. PET changes are seen much earlier then detectable CT changes consistent with imatinib response. PET clearly has an earlier predictive benefit compared to standard CT ((Demetri et al, 2002). In addition CT scanning has a tendency to falsely indicate therapeutic failure or imatinib resistance in cases where myxoid degeneration secondary to anti tyrosine kinase therapy may be observed as an increase in the size of lesions on CT scans. PET scans, on the contrary, show a response in these cases due to decreased metabolic activity (Goldstein et al, 2005).

Tumor resistance to therapy is defined as progression of tumor burden noticed on imaging studies. Response does not always manifest as an immediate decrease in size of tumor lesions but rather as an initial inhibition of growth (Verweij at al, 2004). Absence of any tumor progression after 8-12 weeks of therapy is considered a response to treatment and labeled as stable disease. These patients who respond to the initial therapy should complete of a full 6 month course of treatment of up before surgery is contemplated. Some patients manifest a clear decrease in tumor burden after initiation of therapy, and are labeled as having responsive disease, while an increase in tumor burden (not size alone) despite therapy is defined as resistant disease. According to the Response Evaluation Criteria in Solid tumors (RECIST), initial resistance is defined as objective disease progression within 3 months of randomization (Therasse et al, 2000). This cut off point was selected to include progression documented at first response evaluation after 2 months of therapy but to exclude progression documented at second disease evaluation at 4 months after initiation of therapy. Late resistance was defined as tumor progression noticed after 3 months of initiation of imatinib therapy. The major drawback of the RECIST criteria is the fact that tumor response is judged primarily by unidimensional tumor size. This problem more pronounced in GIST treated with molecularly targeted agents such as imatinib where the tumor can increase in size despite good response. Also, focal progression within a responding GIST can be overlooked with the RECIST criteria (Choi et al, 2007). The Choi criteria use a combination of tumor size and density is better suited in early response evaluation and in predicting long-term prognosis (Benjamin et al, 2007).

Initial resistance was noted in 12% of all cases analyzed by Van Glabbeke and colleagues in 2005. The presence of lung metastasis, absence of liver metastasis and baseline low hemoglobin of less than 8 gm/dl were highly significant adverse prognostic factors, while a high baseline granulocyte count showed borderline significance. A multivariate analysis of patients with late resistance revealed that tumor size, high baseline granulocyte count (> 5 x 10⁹/L), non gastric primary tumor (which would technically include all duodenal GIST’s) and randomization into the low dose group (400 mg /day of imatinib) were independent factors of adverse prognosis. We thus infer that the highest possible tolerable dose of imatinib should be administered to all patients with pancreatoduodenal GIST who require neoadjuvant treatment. Late resistance to imatinib is independently predicted by the size of the lesions present at the commencement of neoadjuvant therapy. Thus, tumor size by itself forms an important prognostic factor for primary disease. Treatment with imatinib is generally safe and well tolerated with the most common adverse effects being anemia, edema, nausea, diarrhea, myalgia, fatigue and skin rash (Fletcher et al, 2002).

There are certain circumstances in which imatinib mesylate may not be useful as a therapeutic agent. Approximately 30 % of GIST’s have been noted to lack KIT gene mutations but instead possess an activating mutation of Platelet Derived Growth Factor Receptor-α (PDGFRα) (Heinrich et al, 2003). Familial GIST with germ line mutations of PDGFRA genes and Neurofibromatosis (NF) genes has been reported. These cases are most likely to present with imatinib resistance due to different tumorigenic mechanisms (Nishida et al, 1998).

M. Defining multimodal therapy: An algorithmic approach

Imatinib mesylate has a role both as neoadjuvant and adjuvant treatment. Based on the literature reviewed so far, the exact indication for imatinib as neoadjuvant therapy in pancreatoduodenal GIST is still controversial. In order to clearly identify the role of imatinib mesylate for treatment of patients with GIST of the duodenum, we created an algorithm based on the salient factors highlighted from limited literature enumerated above and our own experience. Our algorithm (Figure 6) begins at initial presentation, which can be characterized as involving either vague abdominal complaints or upper gastrointestinal bleeding. Each is evaluated as indicated in the algorithm until a diagnosis is established and the staging work-up completed. Most of these cases, as noted, are complex and require a thorough discussion at every institutions Tumor board before a definitive regimen is implemented for each patient. We designed a scoring system based on evaluation of published articles and have incorporated this into the algorithm which would determine those patients who are expected to require neoadjuvant treatment based on tumor size and surgical resectability (Table 3). We quantify tumor size on a scale of 1 to 4, with 1 representing tumors of less than 2 cm in size and 4 assigned to tumors 10 cm or greater in size.

A similar ordinal scale is used to quantify surgical resectability, which is determined by the surgeon’s evaluation of the imaging studies and extent of disease. Surgical resectability is typically categorized into four groups based purely on technical aspects: resectable with no functional deficit; resectable with associated functional deficit; potentially respectable; and unresectable or metastatic disease. Procedures which could potentially result in functional disability include any form of reconstructive procedure other than a wide excision with a primary closure (Billroth II, Whipple’s procedure or a Valdoni-Strong procedure). All such complex procedures receive a score of 2. If the tumor is of questionable resectability based on diagnostic imaging, or if the surgeon is not sure of the surgical plan, we categorize it as potentially resectable and assign a score of 3. Patients with metastatic disease or unresectable disease or those with preoperative biopsy-proven evidence of high grade tumor are assigned a score of 4. We hypothesize that patients with a combined score of 5 or more (potential scores ranging from 2 to 8) require neoadjuvant imatinib therapy.

Imatinib mesylate is indicated in most patients with high grade GIST and grading is usually not possible unless the specimen in resected. In the setting of neoadjuvant therapy, assessment of mitotic index by preoperative biopsy techniques is technically difficult unless an adequate core biopsy specimen is available. At least 50 HPF need to be examined to definitely conclude a low or intermediate grade GIST. We prefer to avoid core needle biopsies due to potential intraperitoneal tumor spillage unless the tumor can be biopsied through an endoscopic approach. In other circumstances, we would proceed with only an FNA, which would complement the CT findings and help in identifying certain high grade GISTs. Although less tissue is sampled with an FNA, the presence of necrosis or higher mitotic figures may help in clinching the diagnosis.

The presence of many mitotic indices in a smaller sample is indicative of a high-grade tumor. FNA specimens are preferred from a pre-operative perspective as they cause less tumor spillage, but they are more challenging with respect to grading the tumor. However, the very presence of mitotic cells and the Ki-67 labeling indexing in FNA specimens have been shown to be significant predictive factors for high grade GIST (Ando et al, 2002).

After the score is calculated, the patients either undergo neoadjuvant therapy or operative exploration. Neoadjuvant therapy is initiated at full dose and patients are reevaluated with a CT and a PET scan at 6 to 8 weeks and again at 16 weeks. Patients with responding or stable disease after the follow up imaging studies undergo the full course of imatinib mesylate. Patients who demonstrate relapse or late resistance are either considered for alternative modalities of treatment with sunitinib or for definitive operative management or debulking procedures depending on the extent of the disease.

Patients who respond to imatinib mesylate therapy are reassessed using the scoring system: if their score is less than 5 they are managed surgically. Those with a score of more than 5 are considered for definitive surgery or debulking procedures depending on the extent of the disease. The aim of operative exploration is to achieve negative margins without tumor spillage and, as indicated in the algorithm, the type of procedure varies depending on tumor location and the size.

Caution has to be exercised when the surgical plan is modified based on therapeutic response to neoadjuvant treatment. Specifically in GIST, although tumor sizes may have decreased, there often is evidence of microscopic disease in portions of the specimen that previously had tumor-related PET activity but are currently PET inactive secondary to the neoadjuvant treatment. Utilization of the Choi criteria helps to assess tumor response in a systematic fashion and is more thorough compared to the classical RECIST criteria (Choi et al, 2004). Based on a study of 36 patients by Goh et al (2006), pathologic response of GIST to imatinib mesylate is usually incomplete and does not correlate with the complete response seen on PET scan. Their findings suggested that surgical resection will continue to play a vital role in the treatment of patients with advanced disease despite a response to imatinib mesylate. Hence we feel that the surgical plan should be decided based on the baseline triphasic CT scan obtained on initial evaluation, and we

Isoform analysis for specific exon 11 point mutation would be helpful in making the diagnosis in such circumstances. Those patients who demonstrate rapidly progressive disease should be considered for sunitinib, and imatinib should be discontinued (Blay et al, 2005). These patients should be considered for any of the ongoing clinical trials, as clear data are still not available for the management of patients with advanced progressive disease, which is uniformly fatal. The role of re-resection is not yet clearly defined. Focal resistance to imatinib mesylate therapy can develop in specific lesions, in which case limited resection of progressing lesions should be considered.

IV. Conclusion

Gastrointestinal stromal tumors might well be called the tumors of the decade due to the tremendous growth in our scientific understanding of them and the controversies still surrounding them. The role of tyrosine kinase inhibitors has been clearly established in halting the progress of the disease in a certain number of patients, and has definitely enhanced the options available for managing these patients. Imatinib has been proven to be a valuable adjunct to surgery. Despite this, surgery remains the gold standard, primary modality for managing these tumors as the option offering the highest chance of cure. Caution has to exercised when dealing with stromal tumors that are CD-117 negative but clinically aggressive. These might be variant forms of GIST. Genetic analysis may be helpful in determining if a KIT gene mutation indeed exists. Endoluminal approaches are safer in establishing pre-operative diagnosis and avoid peritoneal seeding. The essence of surgical management is to avoid violating the tumor capsule and unnecessary manipulation of the tumor itself. We also strongly advocate the principle of obtaining negative margins at the very first specimen. Positive margins at the first specimen are indicative of violation of tumor territory/capsule and negatively impacts prognosis. Hence, we favor a very aggressive surgical approach for managing patients with pancreatoduodenal GIST and dissect well away from the vicinity of the tumor, although this might require major reconstructive operations in certain cases. The scoring system we have designed is based on extrapolating data available in the literature from GIST occurring elsewhere in the GI tract, as the amount of data available on pancreatoduodenal GIST is limited and will take time to accrue. The cut off points in our scoring system are based on clinical decisions pertaining to the cases we have reviewed and experienced. Although arbitrary to a certain degrees, we feel it would serve as a good tool until more data is available which would prompt revision of the scoring system.

Despite limitations due to the paucity of data available, the algorithm we propose could serve as a reasonable guide for managing these aggressive tumors in critical locations such as the pancreatoduodenal complex. Ideally this could serve to streamline the treatment of these patients, and future data when available could be more easily analyzed and outcomes predicted within clinical trials.

References

Ando N, Goto H, Niwa Y, Hirooka Y, Ohmiya N, Nagasaka T, Hayakawa T (2002) The diagnosis of GI stromal tumors with EUS-guided fine needle aspiration with immunohistochemical analysis. Gastrointest Endosc 55, 37-43.

Antonioli DA (1989) Gastrointestinal autonomic nerve tumors. Expanding the spectrum of gastrointestinal stromal tumors. Arch Pathol Lab Med 113, 831-3.

Appelman HD, Helwig EB (1977) Sarcomas of the stomach. Am J Clin Pathol 67, 2-10.

Bagnolo F, Bonassi U, Scelsi R, Testoni PA (1998) Gastric stromal tumour, a rare neoplasm presenting with gastrointestinal bleeding. Eur J Gastroenterol 10, 791-4.

Benjamin RS, Choi H, Macapinlac HA, Burgess MA, Patel SR, Chen LL, Podoloff DA, Charnsangavej C (2007) We should desist using RECIST, at least in GIST. J Clin Oncol 25, 1760-4.

Bernet L, Zuniga A, Cano R (2003) Characterization of GIST/GIPACT tumors by inmunohistochemistry and exon 11 analysis of c-kit by PCR. Rev Esp Enferm Dig 95, 688-91.

Blay JY, Bonvalot S, Casali P, Choi H, Debiec-Richter M, Dei Tos AP, Emile JF, Gronchi A, Hogendoorn PC, Joensuu H, Le Cesne A, McClure J, Maurel J, Nupponen N, Ray-Coquard I, Reichardt P, Sciot R, Stroobants S, van Glabbeke M, van Oosterom A , Demetri GD (2005) GIST consensus meeting panelists. Consensus meeting for the management of gastrointestinal stromal tumors. Report of the GIST Consensus Conference of 20-21 March 2004, under the auspices of ESMO. Ann Oncol 16, 566-78.

Bono P, Krause A, von Mehren M, Heinrich MC, Blanke CD, Dimitrijevic S, Demetri GD, Joensuu H (2004) Serum KIT and KIT ligand levels in patients with gastrointestinal stromal tumors treated with imatinib. Blood 103, 2929-35.

Buchdunger E, Cioffi CL, Law N, Stover D, Ohno-Jones S, Druker BJ, Lydon NB (2000) Abl protein-tyrosine kinase inhibitor STI571 inhibits in vitro signal transduction mediated by c-kit and platelet-derived growth factor receptors. J Pharmacol Exp Ther 295, 139-45.

Carvajal H C, Iturra U S, Justiniano P JC, Bustamante Z M, Contreras P JE, Lombardi S J, Capetillo Fuentes M (2006) Duodenal gastrointestinal stromal tumors. Report to one case. Rev Med Chil 134, 481-4.

Chatzipantelis P, Salla C, Karoumpalis I, Apessou D, Sakellariou S, Doumani I, Papaliodi E, Konstantinou P (2008) Endoscopic ultrasound-guided fine needle aspiration biopsy in the diagnosis of gastrointestinal stromal tumors of the stomach. A study of 17 cases. J Gastrointestin Liver Dis 17, 15-20.

Choi H, Charnsangavej C, de Castro Faria S, Tamm EP, Benjamin RS, Johnson MM, Macapinlac HA, Podoloff DA (2004) CT evaluation of the response of gastrointestinal stromal tumors after imatinib mesylate treatment, a quantitative analysis correlated with FDG PET findings. Am J Roentgenol 183, 1619-28

Choi H, Charnsangavej C, Faria SC, Macapinlac HA, Burgess MA, Patel SR, Chen LL, Podoloff DA, Benjamin RS (2007) Correlation of computed tomography and positron emission tomography in patients with metastatic gastrointestinal stromal tumor treated at a single institution with imatinib mesylate, proposal of new computed tomography response criteria. J Clin Oncol 25, 1753-9.

Contreras-Hernández I, Mould-Quevedo JF, Silva A, Salinas-Escudero G, Villasís-Keever MA, Granados-García V, Dávila-Loaiza G, Petersen JA, Garduño-Espinosa J (2008) A pharmaco-economic analysis of second-lin e treatment with imatinib or sunitinib in patients with advanced gastrointestinal stromal tumours. Br J Cancer 98, 1762-8.

Dalaker K, Harket R (1980) Leiomyoblastoma (epithelioid leiomyomaof the stomach. Acta Chir Scand 146, 141-4.

De Marco G, Roviello F, Marrelli D, De Stefano A, Neri A, Rossi S, Corso G, Rampone B, Nastri G, Pinto E (2005) A clinical case of duodenal gastrointestinal stromal tumor with a peculiarity in the surgical approach. Tumori 91, 261-3.

De Nicola P, Di Bartolomeo N, Francomano F, D'Aulerio A, Innocenti P (2005) Segmental resection of the third and fourth portions of the duodenum after intestinal derotation for a GIST, a case report. Suppl Tumori 4, S108-10.

DeMatteo RP, Lewis JJ, Leung D, Mudan SS, Woodruff JM, Brennan MF (2000) Two hundred gastrointestinal stromal tumors, recurrence patterns and prognostic factors for survival. Ann Surg 231, 51-8.

Demetri GD, van Oosterom AT, Garrett CR, Blackstein ME, Shah MH, Verweij J, McArthur G, Judson IR, Heinrich MC, Morgan JA, Desai J, Fletcher CD, George S, Bello CL, Huang X, Baum CM, Casali PG (2006) Efficacy and safety of sunitinib in patients with advanced gastrointestinal stromal tumour after failure of imatinib, a randomised controlled trial. Lancet 8, 1329-38.

Demetri GD, von Mehren M, Blanke CD, Van den Abbeele AD, Eisenberg B, Roberts PJ, Heinrich MC, Tuveson DA, Singer S, Janicek M, Fletcher JA, Silverman SG, Silberman SL, Capdeville R, Kiese B, Peng B, Dimitrijevic S, Druker BJ, Corless C, Fletcher CD, Joensuu H (2002) Efficacy and safety of imatinib mesylate in advanced gastrointestinal stromal tumors. N Engl J Med 347, 472-80.

Dierkes-Globisch A, Goeller T, Mohr HH (2001) Gastric stromal tumor--a rare cause of an upper gastrointestinal bleeding. Z Gastroenterol 39, 467-70.

Druker BJ, Tamura S, Buchdunger E, Ohno S, Segal GM, Fanning S, Zimmermann J, Lydon NB (1996) Effects of a selective inhibitor of the Abl tyrosine kinase on the growth of Bcr-Abl positive cells. Nat Med 2, 561-6.

Eisenberg BL (2003) Imatinib mesylate, a molecularly targeted therapy for gastrointestinal stromal tumors. Oncology 17, 1615-20.

Elliott DD, Fanning CV, Caraway NP (2006) The utility of fine-needle aspiration in the diagnosis of gastrointestinal stromal tumors, a cytomorphologic and immunohistochemical analysis with emphasis on malignant tumors. Cancer 108, 49-55.

Emory TS, O'Leary TJ (1998) Prognosis and surveillance of gastrointestinal stromal/smooth muscle tumors. Ann Chir Gynaecol 87, 306-10.

Emory TS, Sobin LH, Lukes L, Lee DH, O'Leary TJ (1999) Prognosis of gastrointestinal smooth-muscle (stromaltumors, dependence on anatomic site. Am J Surg Pathol 23, 82-7.

Ernst SI, Hubbs AE, Przygodzki RM, Emory TS, Sobin LH, O'Leary TJ (1998) KIT mutation portends poor prognosis in gastrointestinal stromal/smooth muscle tumors. Lab Invest 78, 1633-6.

Evans HL (1985) Smooth muscle tumors of the gastrointestinal tract. A study of 56 cases followed for a minimum of 10 years. Cancer 56, 2242-50.

Faussone-Pellegrini MS, Pantalone D, Cortesini C (1990) Smooth muscle cells, interstitial cells of Cajal and myenteric plexus interrelationships in the human colon. Acta Anat (Basel139, 31-44.

Fletcher CD, Berman JJ, Corless C, Gorstein F, Lasota J, Longley BJ, Miettinen M, O'Leary TJ, Remotti H, Rubin BP, Shmookler B, Sobin LH, Weiss SW (2002) Diagnosis of gastrointestinal stromal tumors, A Consensus approach Hum Patho 33, 459-65.

Fu K, Eloubeidi MA, Jhala NC, Jhala D, Chhieng DC, Eltoum IE (2002) Diagnosis of gastrointestinal stromal tumor by endoscopic ultrasound-guided fine needle aspiration biopsy--a potential pitfall. Ann Diagn Pathol 6, 294-301.

Goerres GW, Stupp R, Barghouth G, Hany TF, Pestalozzi B, Dizendorf E, Schnyder P, Luthi F, von Schulthess GK, Leyvraz S (2005) The value of PET, CT and in-line PET/CT in patients with gastrointestinal stromal tumours, long-term outcome of treatment with imatinib mesylate. Eur J Nucl Med Mol Imaging 32, 153-62.

Goh BK, Chow PK, Chuah KL, Yap WM, Wong WK (2006) Pathologic, radiologic and PET scan response of gastrointestinal stromal tumors after neoadjuvant treatment with imatinib mesylate. Eur J Surg Oncol 32, 961-3.

Goh BK, Chow PK, Ong HS, Wong WK (2005) Gastrointestinal stromal tumor involving the second and third portion of the duodenum, treatment by partial duodenectomy and Roux-en-Y duodenojejunostomy. J Surg Oncol 91, 273-5.

Gold JS, van der Zwan SM, Gönen M, Maki RG, Singer S, Brennan MF, Antonescu CR, De Matteo RP (2007) Outcome of metastatic GIST in the era before tyrosine kinase inhibitors. Ann Surg Oncol 14, 134-42.

Goldstein D, Tan BS, Rossleigh M, Haindl W, Walker B, Dixon J (2005) Gastrointestinal stromal tumours, correlation of F-FDG gamma camera-based coincidence positron emission tomography with CT for the assessment of treatment response--an AGITG study. Oncology 69, 326-32.

Graadt van Roggen JF, van Velthuysen ML, Hogendoorn PC (2001) The histopathological differential diagnosis of gastrointestinal stromal tumours. J Clin Pathol 54, 96-102.

Heinrich MC, Corless CL, Duensing A, McGreevey L, Chen CJ, Joseph N, Singer S, Griffith DJ, Haley A, Town A, Demetri GD, Fletcher CD, Fletcher JA (2003) PDGFRA activating mutations in gastrointestinal stromal tumors. Science 299, 708-10.

Herrera GA, Pinto de Moraes H, Grizzle WE, Han SG (1984) Malignant small bowel neoplasm of enteric plexus derivation (plexosarcoma. Light and electron microscopic study confirming the origin of the neoplasm. Dig Dis Sci 29, 275-84.

Hinz S, Pauser U, Egberts JH, Schafmayer C, Tepel J, Fandrich F (2006) Audit of a series of 40 gastrointestinal stromal tumour cases. Eur J Surg Oncol 32, 1125-9.

Hirota S, Isozaki K, Moriyama Y, Hashimoto K, Nishida T, Ishiguro S, Kawano K, Hanada M, Kurata A, Takeda M, Muhammad Tunio G, Matsuzawa Y, Kanakura Y, Shinomura Y, Kitamura Y (1998) Gain-of-function mutations of c-kit in human gastrointestinal stromal tumors. Science 279, 577-80.

Hirota S, Isozaki K (2006) Pathology of gastrointestinal stromal tumors. Pathol Int 56, 1-9.

Hompes D, Topal B, Ectors N, Aerts R, Penninckx F (2004) Gastro-intestinal stromal tumour of the duodenum, extreme presentation in two cases. Acta Chir Belg 104, 110-3.

Hughes JA, Cook JV, Said A, Chong SK, Towu E, Reidy J (2004) Gastrointestinal stromal tumour of the duodenum in a 7-year-old boy. Pediatr Radiol 34, 1024-7.

Huizinga JD, Thuneberg L, Kluppel M, Malysz J, Mikkelsen HB, Bernstein A (1995) W/kit gene required for interstitial cells of Cajal and for intestinal pacemaker activity. Nature 373, 347-9.

Huizinga JD, Thuneberg L, Vanderwinden JM, Rumessen JJ (1997) Interstitial cells of Cajal as targets for pharmacological intervention in gastrointestinal motor disorders. Trends Pharmacol Sci 18, 393-403.

Joensuu H, Roberts PJ, Sarlomo-Rikala M, Andersson LC, Tervahartiala P, Tuveson D, Silberman S, Capdeville R, Dimitrijevic S, Druker B, Demetri GD (2001) Effect of the tyrosine kinase inhibitor STI571 in a patient with a metastatic gastrointestinal stromal tumor. N Engl J Med 344, 1052-6.

Joensuu H (2006) Sunitinib for imatinib-resistant GIST. Lancet 368, 1303-4.

Kerr JZ, Hicks MJ, Nuchtern JG, Saldivar V, Heim-Hall J, Shah S, Kelly DR, Cain WS, Chintagumpala MM (1999) Gastrointestinal autonomic nerve tumors in the pediatric population, a report of four cases and a review of the literature. Cancer 85, 220-30.

Kim IH, Kim PS, Lee DH, Choi W, Kim HG, Kim YS, Chu YC (1999) Gastric malignant stromal tumor with long stalk impacted into duodenum. Yonsei Med J 40, 510-3.

Kindblom LG, Remotti HE, Aldenborg F, Meis-Kindblom JM (1998) Gastrointestinal pacemaker cell tumor (GIPACT, gastrointestinal stromal tumors show phenotypic characteristics of the interstitial cells of Cajal. Am J Pathol 152, 1259-69.

Kluppel M, Huizinga JD, Malysz J, Bernstein A (1998) Developmental origin , Kit-dependent development of the interstitial cells of cajal in the mammalian small intestine. Dev Dyn 211, 60-71.

Lasota J, Jasinski M, Sarlomo-Rikala M, Miettinen M (1999) Mutations in exon 11 of c-Kit occur preferentially in malignant versus benign gastrointestinal stromal tumors and do not occur in leiomyomas or leiomyosarcomas. Am J Pathol 154, 53-60.

Lavin P, Hajdu SI, Foote FW Jr (1972) Gastric and extragastric leiomyoblastomas, clinicopathologic study of 44 cases. Cancer 29, 305-11.

Le Tourneau C, Raymond E, Faivre S (2007) Sunitinib, a novel tyrosine kinase inhibitor. A brief review of its therapeutic potential in the treatment of renal carcinoma and gastrointestinal stromal tumors (GIST). Ther Clin Risk Manag 3, 341-8.

Lee CM, Chen HC, Leung TK, Chen YY (2004) Gastrointestinal stromal tumor, Computed tomographic features. World J Gastroenterol 10, 2417-8.

Lee JR, Joshi V, Griffin JW Jr, Lasota J, Miettinen M (2001) Gastrointestinal autonomic nerve tumor, immunohistochemical and molecular identity with gastrointestinal stromal tumor. Am J Surg Pathol 25, 979-87.

Lee JS, Nascimento AG, Farnell MB, Carney JA, Harmsen WS, Ilstrup DM (1995) Epithelioid gastric stromal tumors (leiomyoblastomas, a study of fifty-five cases. Surgery 118, 653-60.

Linton KM, Taylor MB, Radford JA (2006) Response evaluation in gastrointestinal stromal tumours treated with imatinib, misdiagnosis of disease progression on CT due to cystic change in liver metastases. Br J Radiol 79, e40-4.

Manley PW, Cowan-Jacob SW, Buchdunger E, Fabbro D, Fendrich G, Furet P, Meyer T, Zimmermann J (2002) Imatinib, a selective tyrosine kinase inhibitor. Eur J Cancer 38, S19-27.

Mazur MT, Clark HB (1983) Gastric stromal tumors. Reappraisal of histogenesis. Am J Surg Pathol 7, 507-19.

Meesters B, Pauwels PA, Pijnenburg AM, Vlasveld LT, Repelaer van Driel OJ (1998) Metastasis in a benign duodenal stromal tumour. Eur J Surg Oncol 24, 334-5.

Miettinen M, Lasota J (2006) Gastrointestinal stromal tumors, review on morphology, molecular pathology, prognosis, and differential diagnosis. Arch Pathol Lab Med 130, 1466-78.

Miettinen M, Virolainen M, Maarit-Sarlomo-Rikala (1995) Gastrointestinal stromal tumors--value of CD34 antigen in their identification and separation from true leiomyomas and schwannomas. Am J Surg Pathol 19, 207-16.

Miettinen M.Gastrointestinal stromal tumors (1988) An immunohistochemical study of cellular differentiation. Am J Clin Pathol 89, 601-10.

Neuhaus SJ, Clark MA, Hayes AJ, Thomas JM, Judson I (2005) Surgery for gastrointestinal stromal tumour in the post-imatinib era. ANZ J Surg 75, 165-72.

Nishida T, Hirota S, Taniguchi M, Hashimoto K, Isozaki K, Nakamura H, Kanakura Y, Tanaka T, Takabayashi A, Matsuda H, Kitamura Y (1998) Familial gastrointestinal stromal tumours with germline mutation of the KIT gene. Nat Genet 19, 323-4.

Pauwels P, Debiec-Rychter M, Stul M, De Wever I, Van Oosterom AT, Sciot R (2005) Changing phenotype of gastrointestinal stromal tumours under imatinib mesylate treatment, a potential diagnostic pitfall. Histopathology 47, 41-7.

Pierie JP, Choudry U, Muzikansky A, Yeap BY, Souba WW, Ott MJ (2001) The effect of surgery and grade on outcome of gastrointestinal stromal tumors. Arch Surg 136, 383-9.

Ranchod M, Kempson RL (1977) Smooth muscle tumors of the gastrointestinal tract and retroperitoneum, a pathologic analysis of 100 cases. Cancer 39, 255-62.

Raut CP, Posner M, Desai J, Morgan JA, George S, Zahrieh D, Fletcher CD, Demetri GD, Bertagnolli MM (2006) Surgical management of advanced gastrointestinal stromal tumors after treatment with targeted systemic therapy using kinase inhibitors. J Clin Oncol 24, 2325-31.

Rossi CR, Mocellin S, Mencarelli R, Foletto M, Pilati P, Nitti D, Lise M (2003) Gastrointestinal stromal tumors, from a surgical to a molecular approach. Int J Cancer 107, 171-6.

Rudolph P, Chiaravalli AM, Pauser U, Oschlies I, Hillemanns M, Gobbo M, Marichal M, Eusebi V, Hofler H, Capella C, Kloppel G (2002) Gastrointestinal mesenchymal tumors -immunophenotypic classification and survival analysis. Virchows Arch 441, 238-48.

Rumessen JJ, Mikkelsen HB, Thuneberg L (1992) Ultrastructure of interstitial cells of Cajal associated with deep muscular plexus of human small intestine. Gastroenterology 102, 56-68.

Rumessen JJ (1996) Ultrastructure of interstitial cells of Cajal at the colonic submuscular border in patients with ulcerative colitis. Gastroenterology 111, 1447-55.

Rutkowski P, Nowecki Z, Nyckowski P, Dziewirski W, Grzesiakowska U, Nasierowska-Guttmejer A, Krawczyk M, Ruka W (2006) Surgical treatment of patients with initially inoperable and/or metastatic gastrointestinal stromal tumors (GISTduring therapy with imatinib mesylate. J Surg Oncol 93, 304-11.

Rutkowski P, Symonides M, Zdzienicki M, Siedlecki JA (2008) Developments in targeted therapy of advanced gastrointestinal stromal tumors. Recent Patents Anticancer Drug Discov 3, 88-99.

Sakamoto Y, Yamamoto J, Takahashi H, Kokudo N, Yamaguchi T, Muto T, Makuuchi M (2003) Segmental resection of the third portion of the duodenum for a gastrointestinal stromal tumor, a case report. Jpn J Clin Oncol 33, 364-6.

Sarlomo-Rikala M, Kovatich AJ, Barusevicius A, Miettinen M (1998) CD117, a sensitive marker for gastrointestinal stromal tumors that is more specific than CD34. Mod Pathol 11, 728-34.

Saul SH, Rast ML, Brooks JJ (1987) The immunohistochemistry of gastrointestinal stromal tumors.Evidence supporting an origin from smooth muscle. Am J Surg Pathol 11, 464-73.

Sawaki A, Ohashi K, Yamao K, Inada K, Shimizu Y, Matsuura A, Nakamura T, Suzuki T, Hara K, Okubo K, Ohno R (2003) Effect of a tyrosine kinase inhibitor STI571 in a patient with hepatic metastases from a duodenal gastrointestinal stromal tumor. J Gastroenterol 38, 690-4.

Schubert ML, Moghimi R (2006) Gastrointestinal stromal tumor (GIST. Curr Treat Options Gastroenterol 9, 181-8.

Segal A, Carello S, Caterina P, Papadimitriou JM, Spagnolo DV (1994) Gastrointestinal autonomic nerve tumors, a clinicopathological, immunohistochemical and ultrastructural study of 10 cases. Pathology 26, 439-47.

Shankar S, vanSonnenberg E, Desai J, Dipiro PJ, Van Den Abbeele A, Demetri GD (2005) Gastrointestinal stromal tumor, new nodule-within-a-mass pattern of recurrence after partial response to imatinib mesylate. Radiology 235, 892-8.

Singer S, Rubin BP, Lux ML, Chen CJ, Demetri GD, Fletcher CD, Fletcher JA (2002) Prognostic value of KIT mutation type, mitotic activity, and histologic subtype in gastrointestinal stromal tumors. J Clin Oncol 20, 3898-905.

Sircar K, Hewlett BR, Huizinga JD, Chorneyko K, Berezin I, Riddell RH (1999) Interstitial cells of Cajal as precursors of gastrointestinal stromal tumors. Am J Surg Pathol 23, 377-89.

Stratopoulos C, Soonawalla Z, Piris J, Friend PJ (2006) Hepatopancreatoduodenectomy for metastatic duodenal gastrointestinal stromal tumor. Hepatobiliary Pancreat Dis Int 5, 147-50.

Taniguchi M, Nishida T, Hirota S, Isozaki K, Ito T, Nomura T, Matsuda H, Kitamura Y (1999) Effect of c-kit mutation on prognosis of gastrointestinal stromal tumors. Cancer Res 59, 4297-300.

Therasse P, Arbuck SG, Eisenhauer EA, Wanders J, Kaplan RS, Rubinstein L, Verweij J, Van Glabbeke M, van Oosterom AT, Christian MC, Gwyther SG (2000) New guidelines to evaluate the response to treatment in solid tumors. European Organization for Research and Treatment of Cancer, National Cancer Institute of the United States, National Cancer Institute of Canada. J Natl Cancer Inst 92, 205-16.

Torihashi S, Nishi K, Tokutomi Y, Nishi T, Ward S, Sanders KM (1999) Blockade of kit signaling induces transdifferentiation of interstitial cells of cajal to a smooth muscle phenotype. Gastroenterology 117, 140-8.

Tornoczky T, Kalman E, Hegedus G, Horvath OP, Sapi Z, Antal L, Jakso P, Pajor L (1999) High mitotic index associated with poor prognosis in gastrointestinal autonomic nerve tumour. Histopathology 35, 121-8.

Tortella BJ, Matthews JB, Antonioli DA, Dvorak AM, Silen W (1987) Gastric autonomic nerve (GANtumor and extra-adrenal paraganglioma in Carney's triad. A common origin. Ann Surg 205, 221-5.

Tuveson DA, Willis NA, Jacks T, Griffin JD, Singer S, Fletcher CD, Fletcher JA, Demetri GD (2001) STI571 inactivation of the gastrointestinal stromal tumor c-KIT oncoprotein, biological and clinical implications. Oncogene 20, 5054-8.

van de Rijn M, Hendrickson MR, Rouse RV (1994) CD34 expression by gastrointestinal tract stromal tumors. Hum Pathol 25, 766-71.

Van Glabbeke M, Verweij J, Casali PG, Le Cesne A, Hohenberger P, Ray-Coquard I, Schlemmer M, van Oosterom AT, Goldstein D, Sciot R, Hogendoorn PC, Brown M, Bertulli R, Judson IR (2005) Initial and late resistance to imatinib in advanced gastrointestinal stromal tumors are predicted by different prognostic factors, a European Organisationfor Research and Treatment of Cancer-Italian Sarcoma Group-Australasian Gastrointestinal Trials Group study. J Clin Oncol 23, 5795-804.

Winfield RD, Hochwald SN, Vogel SB, Hemming AW, Liu C, Cance WG, GrobmyerSR (2006) Presentation and management of gastrointestinal stromal tumors of the duodenum. Am Surg 72, 719-22.

Yamada T, Ichikawa H (1974) X-ray diagnosis of elevated lesions of the stomach. Radiology 110, 79-83.

Yeh CN, Chen TW, Liu FY, Jan YY, Chen MF (2006) Genetic changes in advanced gastrointestinal stromal tumor (GISTpatients during imatinib mesylate treatment. Langenbecks Arch Surg 391, 615-21.

Raja R Gopaldas

	Case I	Case II	Case III
Age at presentation(yrs)	32	49	52
Gender	Male	Male	Male
Mode of presentation	Anemia	Abdominal pain	Abdominal pain
Hemoglobin ( g/dL)	< 4	8.1	7.4
Procedure	Whipple	Whipple	Antrectomy/DI resection/Billroth II
Duodenum involved	D2	D3	D1
Follow up	3 yrs	6 yrs	2 yrs
Recurrence	No	No	No
Imatinib(adjuvant)	No	No	Yes
Pathology
Tumor Size	8 X 5 X 4 cm	4 X 3 X 3 cm	5 X 4 X 3 cm
Chromogranin A	Negative	Negative	Negative
Synaptophysin	Positive	Negative	Positive
CD 117	Positive	Positive	Positive
CD 34	Positive	Positive	Negative
Neuron specific enolase	Positive	Negative	Positive
CAM 5.2/AE-1 Keratins 39, 40, 43, 48, 50, 50.6 kD	Negative	Negative	Negative
S-100 Protein	Positive	Negative	Positive
MART-1 Melanoma associated marker	Negative	Negative	Negative
Vimentin	Positive	Positive	Positive
Mitoses /50HPF	1	<1	7

GIST (Gastrointestinal stromal tumors)	CD -117 ( c-kit) positive
Leiomyoblastoma	Predominant spindle myoid cells
Plexosarcoma (GANT)	Predominant epithelioid cells
GINST (Gastrointestinal negative stromal tumors)	CD -117 ( c-kit) negative CD - 34 positive
GILT (Gastrointestinal leiomyogenic tumors)	CD -117 ( c-kit) negative α- Smooth muscle actin or desmin
GIGT (Gatrointestinal glial/schwannian tumors)	CD -117 ( c-kit) negative S-100 or glial fibrillary acidic protein
GIFT (Gastrointestinal fibrous tumors)	CD -117 ( c-kit) negative vimentin
GINT (Gastrointestinal neuronal /glial tumors)	CD -117 ( c-kit) negative Neuronal /glial markers

Tumor Size		Surgical resectability		Total score
cm	Score		Score	Total score
≤2	1	Clearly Resectable with no functional deficit	1	· Score is a summative score obtained by adding the scores. · Score of less than 5 does not require neoadjuvant therapy · Score equal to or more than 5 requires neoadjuvant therapy with imatinib mesylate · Neoadjuvant therapy may convert unresectable tumors to resectable ones, but may not necessarily assure negative surgical margin · Any surgery involving a reconstruction of the duodenum other than a wide excision with primary closure is considered as functional deficit
2< tumor ≤ 5	2	Clearly Resectable with associated functional deficit	2
5< tumor ≤ 10	3	Potentially resectable	3
Tumor > 10	4	Unresectable Or High Grade on pre-operative biopsy	4

Tumor Size		Tumor Grade		Surgical outcome	Total score
cm	Score	cm	Score	Multiplier Score	{Size score + Grade score} x { multiplier} Score less than 4 does not require adjuvant therapy Score equal to 4 or more requires post operative adjuvant therapy All patients who received neoadjuvant therapy must receive post op therapy irrespective of score.
≤2	1	Low	1	No residual disease X 1
2< tumor ≤ 5	2	Intermediate	2	No residual disease X 1
5< tumor ≤ 10	3	High	3	Residual disease X 2
Tumor > 10	4			Residual disease X 2

Research Article

Received: 22 May 2008; Revised: 20 June 2008

Accepted: 30 June 2008; electronically published: September 2008

II. Case description

i. Pathology

i. Pathology

A. Defining GIST

E. Anatomic distribution

F. Clinical presentation

G. Diagnostic and staging workup

H. Surgical management

I. Laparoscopy

J. The role of imatinib mesylate

M. Defining multimodal therapy: An algorithmic approach

N. Adjuvant therapy

References