First reported in 1998, most GISTs are characterized by an oncogenic mutation in the Kit tyrosine kinase (CD 117), allowing spontaneous (ligand-independent) receptor dimerization and kinase activation [13, 14]. The c-Kit expression distinguishes GISTs from tumors of smooth muscle cells. GISTs are the most common non-epithelial tumors of the GI tract. The diagnosis of GIST has dramatically increased since 1992, and survival has dramatically improved since 2002. Size and mitotic count are the most prognostic features and are used for risk stratification . All tumors larger than 2 cm have a risk of recurrence and tumor exceeding 5 cm should be considered potentially malignant.
Initial treatment of localized GIST should aim at complete resection of the tumor with margins of 1–2 cm. Segmental resection of small bowel or colon is adequate, no lymphatic dissection is required. Small tumors (< 3 cm) of the stomach could be excised by a laparoscopic approach . Larger tumors request functional gastric resection like antrectomy or resection of the gastric fundus with tube forming . Locoregionally advanced tumors or those poorly positioned requiring total gastrectomy or other extended resection should be considered for neoadjuvant treatment with imatinib mesylate and afterwards re-evaluated for resection [18, 19]. The response rate to be expected from drug therapy is 75% to 80%. It is reasonable to consider the disease as initially as "unresectable" without incurring risk of unacceptable morbidity or functional deficit and therefore to use imatinib mesylate therapy as the first-line anticancer therapy.
In our case, primary resection would have necessitated a multivisceral resection including the distal part of the esophagus, the proximal stomach, spleen and a part of the diaphragm to remove the tumor without contamination and clear margins. After pre-treatment with imatinib mesylate and relevant tumor shrinkage segmental resection was possible. Reconstruction of the upper part of the GI tract after resection of carcinoma of the GEJ usually necessitates esophago-gastrostomy or a Roux-en-Y procedure. Both are often followed by considerable postoperative morbidity mainly due to acidic or biliary esophageal reflux. Dumping syndrome and weight loss are sequelae of excluding the duodenal passage . For locally advanced carcinomas of the esophagus and the GEJ requiring extended lymph node dissection this procedural specific morbidity and mortality is accepted.
Unlike, the resection of GIST tumors does not require lymph node dissection. For these reasons GIST of the GEJ should be treated by limited resection and optimal functional reconstruction if the size of the tumor allow. In 1955, Merendino and Dillard published a technique to reconstruct the esophagogastric passage and to prevent reflux and esophagitis following resection of the GEJ . Initially, the procedure was developed for a variety of benign diseases like severe esophagitis, stricture or cardiospasm. In its final version, the operation consists of the interposition of an isoperistaltic segment of jejunum between the esophagus and stomach, with bilateral vagotomy and a pyloroplasty. In a series of patients who underwent this operation the mean Gastrointestinal Quality of Life Index did not differ from that of healthy controls . We regard the Merendino procedure as an ideal indication for resectable GIST of the proximal stomach or GEJ. In our case, the clinical response to imatinib mesylate with tumor shrinkage prevented an abdomino-thoracic approach with multivisceral resection and allowed a reconstruction of a functional competent GEJ.
Neoadjuvant treatment with imatinib mesylate for locally advanced GIST represents a not yet fully established strategy to handle large GIST that can be resected only with curative intent by major procedures accompanied with organ function loss, i.e. Whipple' procedure for GIST of the duodenum or abdomino-perineal excision for GIST of the rectum or recto-vaginal septum. Two phase II trials currently explore this option in a standardized fashion. The study of the RTOG S-0132 initially used 8 weeks (now 3 months) treatment with imatinib mesylate at 400 mg daily upfront to resection. The so-called Apollon study (CSTI571 BDE43) foresees 6 months of pre-treatment. It is still matter of debate whether treatment with imatinib mesylate should be continued postoperatively. The experience of several centers [, BFR14 study] as well as the guidelines of the NCCN recommend a minimum treatment period with imatinib mesylate of 12 months, thus we continued therapy after recovery in our patient. Response to treatment fulfilling the criteria of a partial remission according to RECIST criteria  requires 3–6 months of therapy . Therefore special criteria of contrast media uptake in CT or MRI have been established [26, 27]. Positron emission tomography with 18F-fluoro-desoxyglucose (FDG) is an ideal tool to monitor treatment effects as it demonstrates shut-down of the tumor metabolism as early as 24 h or 72 hours . In case CT does not show tumor shrinkage early, treatment can be continued safely if PET documents stop of proliferative activity. Also in our patient, 18F-FDG PET was antecedent to CT in demonstrating response to imatinib mesylate and allowed us to complete the full course of preoperative therapy. PET otherwise has been used successfully in epithelial cancer of the esophagus to evaluate treatment with preoperative radiochemotherapy .