The purpose of this study is to determine whether the newly constructed CE (Conformité Européenne) marked drill biopsy instrument for flexible endoscopes can make possible representative biopsies of the increasing numbers of submucosal tumours in the gastrointestinal tract. The amount of submucous tissue ( micrograms and in % of all the biopsy) harvested with the new drill biopsy instrument used in flexible endoscopy investigations is measured in the study and compared to the amount of submucous tissue harvested with the existing biopsy forceps.

Background: With the existing biopsy instruments for flexible endoscopy it's only possible to achieve superficial biopsies comprising the mucosa and very seldom some small fragments of the submucosa. It is, however, increasingly important to get biopsies from the submucous tissues because the tumors that are on the increase more than any other in the stomach is the GIST (GastroIntestinal Stroma Tumours) that grows in the submucosa. Cytology is not enough for a reliable diagnosis and to often not representative. Furthermore there is a well-known risk for tract seeding when sampling through the abdominal wall with a fine needle instrument.

The investigators have invented a new biopsy instrument for flexible endoscopy that is a drill within a cylinder that can be used for biopsies of the submucous tissues, for instance GIST-tumours, see www.BIBBInstruments.com.

Hypothesis: Will it be possible to achieve representative submucous tissue from the gastrointestinal tract with the newly constructed drill biopsy instrument for flexible endoscopy. The investigators hypothesize that the amount of tissue harvested with the new drill biopsy instrument will give significantly more amount of tissue to evaluate for the pathologist than the existing biopsy forceps.

Method: Ten patients with tumours in the submucosa of the stomach are examined with a gastroscope. The investigator decides where the first biopsy should be taken and randomly the investigator blindly get the new drill instrument or the conventional biopsy forceps and take the biopsy on the decided localization. The next site of the tumour is decided and again blindly the investigator get one of the two biopsy instruments according to a previously decided scheme. Six biopsies (three with each of the two biopsy tools) in every patient are harvested in ten patients. With altogether 30 specimen for each biopsy instrument and a difference of 1/3 which is a low calculated difference 10 patients as above will give a 80 % Power (Chi-2 = 5.40, p = 0.02*). The amount of submucous tissue harvested with the two biopsy instruments are measured in microgram and as percentage of all the biopsy and compared. The quality of the specimens are also calculated.

Inclusion Criteria:

• Submucous gastrointestinal tumours

Exclusion Criteria:

• No gastrointestinal submucous tumor
• Submucosal hemangioma