P-0143 - Gastrointestinal stromal tumours in a cohort of South African patients

Date 28 June 2014
Event World GI 2014
Session Poster Session
Topics GIST
Cancer Aetiology, Epidemiology, Prevention
Presenter Barbara Robertson
Citation Annals of Oncology (2014) 25 (suppl_2): ii14-ii104. 10.1093/annonc/mdu165
Authors B. Robertson1, R. Ramesar2, A. Hunter3, M. Bernon1, M. Locketz4, E. Panieri1, G. Baker5, P. Willem5, A. Vorster2
  • 1Groote Schuur Hospital, University of Cape Town, Cape Town/ZA
  • 2MRC Human Genetics Research Unit, University of Cape Town, Cape Town/ZA
  • 3University of Cape Town, Groote Schuur Hospital, Cape Town/ZA
  • 4National Health Laboratory Services, University of Cape Town, Cape Town/ZA
  • 5National Health Laboratory Services, University of Witwatersrand, Johannesburg/ZA



Gastrointestinal stromal tumours (GISTs) have not been extensively studied in African populations, which are known to possess great genetic diversity. As mutational status has been linked with clinical response to imatinib and altered mutational profiles may affect disease characteristics, it is important to document the disease profile in African patients. The aim of this study was to review a cohort of South African patients and assess patient and tumour characteristics and treatment outcome.


All patients with GISTs, referred to Groote Schuur Hospital during 2004 to 2013, were reviewed. Patient demographics, tumour characteristics, treatment and treatment outcomes were recorded. In addition, mutational analysis was performed on formalin fixed paraffin embedded tissue sections.


There were 69 patients in the cohort, with an even gender distribution and a median age of 56 years. Three patients had neurofibromatosis type 1. The most common primary tumour sites were the stomach (56%) and small bowel (23%). Further primary sites included the rectum (6%), mesentery (4%), anus, oesophagus and peritoneum (1.5% each). The primary site could not be established in 6% of patients. The largest tumour diameter had a median of 105 mm. For 39% of the patients the mitotic count was >5/50 high power fields. The primary tumour was resected in 43 patients. A total of 40 patients received imatinib. Five required treatment prior to surgery, four received post-operative adjuvant imatinib while the remaining patients received imatinib for either residual disease after surgery, metastatic disease or locally advanced unresectable disease. For the 36 patients treated with imatinib for locally advanced or metastatic disease, 86% had either a partial response or prolonged stable disease. The median progression free survival was 21.6 months and the median overall survival was 37.5 months. Second line treatment with sunitinib was available for only two patients. Mutational analysis was performed on 34 patients. The distribution of mutations was as follows: KIT exon 11: 67.6%, KIT exon 9: 2.9%, KIT exon 17: 5.8% and PDGFRA exon 18: 5.8% while 20.5% were wild type. Interestingly, one patient had mutations in PDGFRA exon 12 as well as KIT exon 17. Of the three patients who did not respond to imatinib, one was wild type and the mutational status of the others could not be determined.


In this study it was noted that the median age was younger than expected and there was a higher percentage of patients with wild type GISTs than for previously reported populations. Only one patient had a KIT exon 9 mutation. While mutations in KIT and PDGFRA are generally thought to be mutually exclusive, one patient had mutations in both genes. Despite these observations, the distribution of primary sites and the response to imatinib was consistent with that in the literature. Although this is a small sample size, the pattern of mutations of Sotuh African patients appears to be distinct.