An audit of surgical management of gastrointestinal stromal tumours (GIST).

AIM: To analyze GIST outcome after primary resection and to determine if a new grading system could adequately predict there prognosis.

METHODS: A retrospective review (1993-2002) identified 80 patients who underwent primary surgical resection for, c-KIT positive, GIST. Follow-up was complete for all patients (median follow-up 42, range 1-132, months). GIST were classified as low or high grade according to the following parameters: size, mitotic rate, mitotic index (MiB1), presence of necrosis, invasion of adjacent structure and presence of metastasis.

RESULTS: GIST originated from the stomach (46), small bowel (30), colon and rectum two and mesentery two. At surgery, 94% of cases presented with localized disease and 6% blood born metastasis with or without lymph node invasion. Resections were complete (R0) in 72 cases. R0 resection correlated with prognosis (p<0.01). Sixty GIST were classified as low grade (median follow-up 60 months) and 20 as high grade (median follow-up 27 months). Five-year actuarial survival of patients with low or high grade GIST were of 95 and 21%, respectively, (p<0.001).

CONCLUSION: Prognosis of GIST after surgical treatment is influenced by completeness of primary resection and tumour malignant potential. Low grade GIST have an excellent prognosis after surgery alone, while high grade GIST have a high rate of recurrence after primary resection. Adjuvant treatment should be advocated for patient with either high grade GIST or after incomplete primary resection. The presented grading system can reliably predict GIST outcome after primary surgical treatment. Complete surgical resection offers good chance of cure for low grade GIST, while for high grade GIST surgery alone is not sufficient. The presented grading system could be used to identify patients who may benefit of adjuvant treatment with imatinib mesylate after GIST resection.