The risk criteria for assessing the natural course of primary GISTs were validated, but additional independent prognostic factors-primary tumor location and sex--were also identified.

Purpose: Although the mutational status in gastrointestinal stromal tumors (GIST) can predict the response to treatment with tyrosine kinase inhibitors, the role of tumor genotype as a prognostic factor remains controversial. The ConticaGIST study sought to determine the pathologic and molecular factors associated with disease-free survival (DFS) in patients with operable, imatinib-naive GIST.Experimental Design: Clinicopathologic and molecular data from 1,056 patients with localized GIST who underwent surgery with curative intention (R0/R1) and were registered in the European ConticaGIST database were prospectively obtained and reviewed. Risk of tumor recurrence was stratified using the modified NIH criteria. The median follow-up was 52 months.Results: On testing for potential prognostic parameters, the following were associated with inferior DFS on multivariable Cox model analysis: primary nongastric site, size >10 cm, mitotic index >10 mitoses per 50 high power field, and the KIT exon 9 duplication [hazard ratio (HR), 1.47; 95% confidence interval (CI), 0.9-2.5; P ¼ 0.037] and KIT exon 11 deletions involving codons 557 and/or 558 [KITdel-inc557/558; HR, 1.45; 95% CI, 1.0-2.2; P ¼ 0.004]. Conversely, PDGFRA exon 18 mutations were indicators of better prognosis [HR, 0.23; 95% CI, 0.1-0.6; P ¼ 0.002]. KITdel-inc557/558 were an adverse indicator only in GIST localized in the stomach (P < 0.001) but not in tumors with nongastric origin. In gastric GIST, all other mutations presented remarkably superior 5-year DFS.Conclusions: In conclusion, tumor genotype is an independent molecular prognostic variable associated with gastric GIST and should be used for optimizing tailored adjuvant imatinib treatment.

Gastrointestinal stromal tumors (GISTs) comprise a biologically diverse group of neoplasms with respect to activating mutations in either KIT or PDGFRA, histology, anatomical site of origin, and clinical aggressiveness. In this study, we applied the high resolution array-based comparative genomic hybridization (array-CGH) technology to 66 primary GISTs (40 gastric and 26 nongastric, 48 with KIT and 18 with PDGFRA mutations) for identification of novel high-level alterations and for characterization of genotype-related genomic changes. All cases had genomic imbalances with the highest occurrence of 14q (73%), 1p (62%), 22q (59%), 15q (38%), and 13q (29%) losses. Our data indicate that loss of chromosome 14 and/or 22 is an early change in GIST tumorigenesis irrespective of tumor genotype. Furthermore, DNA copy number changes showed a site dependent pattern. These included lower incidence of losses at 14q (87% vs. 35%), and higher frequency of losses at 1p (45% vs. 85%) and 15q (17% vs. 69%) in nongastric versus gastric site (P<0.001 for all). However, in the multivariate analysis with adjustment to tumor risk stratification, only the 14q loss site-dependent pattern of distribution retained its significance. These findings suggest that loss of 14q is a relatively less frequent genetic event in the development of nongastric GISTs, the lack of which is most likely substituted by the accumulation of 1p/15q and other changes. The novel minimal overlapping regions of deletion at 1p (1p36.32-1p35.2, 1p34.1, and 1p22.1-1p21.3), 13q (13q14.11-q14.2 and 13q32.3-q33.1), and 15q23 were delineated, which point to chromosomal regions that may harbor genes relevant to the development of these neoplasms.

Clear cell sarcoma (CCS) is a rare malignant soft tissue tumor particularly associated with tendons and aponeuroses. The cytogenetic hallmark is the translocation t(12;22)(q13; q12) resulting in a chimeric EWS/ATF1 gene in which the 3 -terminal part of EWS at 22q is replaced by the 3 -terminal part of ATF1 at 12q. To date, only 13 cases of CCS have been analyzed for fusion genes at the transcription level, and there is no information about the breakpoints at the genomic level. In the present study, we describe the molecular genetic characteristics of CCS from 10 patients. Karyotypes were obtained from 10 cases, 7 of which showed the characteristic t(12;22). As an initial step in the characterization of the EWS/ATF1 and ATF1/EWS chimeras, we constructed an exon/ intron map of the ATF1 gene. The entire ATF1 gene spanned >40 kb and was composed of 7 exons. Intron 3, in which most of the genomic breakpoints occurred, was to a large extent (83%) composed of repetitive elements. RT-PCR amplified EWS/ATF1 cDNA fragments in all patients and ATF1/EWS cDNA fragments in 6 of 10 patients. Four types of EWS/ATF1 chimeric transcript, designated types 1-4, were identified. The most frequent chimeric transcript (type 1) was an inframe fusion of exon 8 of EWS with exon 4 of ATF1. This was the only chimeric transcript in 5 patients but found together with other variants in 3 tumors. The type 2 transcript of EWS/ATF1, an in-frame fusion of exon 7 of EWS with exon 5 of ATF1, was detected in 4 patients, as the only transcript in 1 case and together with other variants in 3 cases. An in-frame fusion of exon 10 of EWS with exon 5 of ATF1 (type 3) was found in 1 patient as the only transcript, and an out-of-frame fusion of EWS exon 7 with ATF1 exon 7 (type 4) was detected in 1 patient together with type 1 and type 2 transcripts. Sequencing of the amplified ATF1/EWS cDNA fragments showed in 5 patients that ATF1 exon 3 was fused with EWS exon 10, resulting in an out-of-frame chimeric transcript. In 1 case, nt 428 of ATF1 (exon 4) was fused with EWS exon 8; at the junction, there was an insertion of 4 nucleotides, also resulting in an out-of-frame transcript. Genomic extra long PCR and sequence analysis mapped the genomic breakpoints to introns 7, 8 and 9 of EWS and intron 3 and exon 4 of ATF1. While a simple end-to-end fusion was observed in 2 cases, additional nucleotides were found at the junctions in 2 other cases. In addition, topoisomerase I consensus sequences were found close to the junctions, suggesting that this enzyme may participate in the genesis of the EWS/ATF1 fusion.

KIT/PDGFRA mutational status has prognostic significance for patients' outcome and may help in management of patients with GISTs.