Because the application of liquid biopsy in GIST has only been reported in the last year, few but promising data are now available [
21,
22]. The feasibility of liquid biopsy and its application in GISTs was reported for the first time at the 2013 ASCO Annual Meeting [
21,
22]. Using both Sanger sequencing on tumor tissue and Beads Emulsion Amplification Magnetics technology on plasma, Demetri
et al. have analyzed the kinase genotypes in a subgroup of GIST patients in a phase III GRID trial [
21]. They found 84% concordance between plasma and tissue for detection of primary
KIT mutations. Additionally, they more easily detected secondary mutations in plasma (47%) than in tissue (12%). In contrast, the assay was less sensitive for the detection of primary
KIT exon 11 mutations in plasma DNA (12% vs 43%), this apparent discrepancy being partly attributable to the study design, which was specifically targeted to secondary mutations, but mostly to the extensive heterogeneity of primary
KIT exon 11 mutations that hampers the development of specific assays for each possible mutation carried by the tumor. These findings highlight a possible limitation of liquid biopsy that is shared by all other technologies specifically designed to identify
KIT/PDGFRA mutations in GIST management; namely, the wide variability of tyrosine kinase mutations. Thus, less targeted approaches may be better able to provide a complete sequence of the hotspot exons such as CAPP-Seq or TAm-Seq. However, even in this scenario the very large insertions and deletions barely detected by massively parallel sequencing would constitute a drawback [
23,
24]. Maier
et al. retrospectively used 25 different allele-specific L-polymerase chain reaction assays covering
KIT and
PDGFRA mutations to examine 291 plasma samples from 38 GIST patients and correlated the detection of mutated ctDNA with disease status (active disease vs complete response [CR]) [
22]. Interestingly, they found that, in the clinical setting, more patients with active disease, defined as patients having at least one lesion progressing or responding to treatment, had positive results than did CR patients without evidence of residual disease after surgery. They also examined whether the amount of ctDNA correlated with tumor radiological response and reported repeated positive test results and increasing mutant ctDNA in patients with disease progression, negative to positive conversions in patients with relapse, and positive to negative conversions in patients responding to TKIs. Taken together, these findings indicate that mutant ctDNA can be detected and quantified in the plasma of GIST patients and, notably, that the amount of mutant ctDNA correlates with tumor response, suggesting that this approach is feasible and can be used as a surrogate biomarker for predicting both tumor response and relapse in GIST patients.
Although very few data have been accumulated to date, liquid biopsy seems to be a promising tool in GIST management, offering a wide spectrum of clinical applications. The well-known kinase genotype and its relationship with tumor response to TKIs has made GIST a paradigm of molecular biology among solid tumors and provides a strong rationale for applying liquid biopsy in this disease. In particular, the possibility of taking serial blood samples to assess real-time molecular modifications during the course of treatment may identify the development of heterogeneous resistant clones, thereby optimizing the timing of changes in therapeutic strategy. However, because the therapeutic armamentarium available for advanced GIST is limited, this noninvasive approach should not be applied only to anticipate changes in treatment, but also to guide clinicians in re-challenging with drugs according to a rotation strategy.
In addition to offering a dynamic picture of molecular changes during the course of a disease that could serve as early biomarkers of tumor response to TKIs, other intriguing applications of liquid biopsy in GIST management are possible. First, it could be used to assess minimal residual disease after radical resection of a primary tumor, thus accurately identifying patients at high risk of recurrence for whom adjuvant treatment is indicated. Plainly, the main limitation of this application is the ability of the methods used to detect very small amounts of ctDNA; more sensitive assays should be developed for this purpose. Moreover, in the light of the extreme biological heterogeneity of GIST, especially those without the known KIT/PDGFRA or BRAF/RAS/NF1 or SDH mutations (termed WT GIST), liquid biopsy may be the best tool for obtaining a wide molecular picture of GIST by using whole genome analyses with high-throughput technologies, such as next-generation sequencing approaches. ctDNA could be used as a source for identifying novel chromosomal adaptations such as rearrangements and translocations, thus overcoming the limitations of creating fresh tissue banks. This would facilitate a more detailed molecular classification of this disease and shed light on new diagnostic and prognostic biomarkers and eventually new potential targets, thereby extending the therapeutic armamentarium for a subset of GIST. Finally, liquid biopsy could also be used to select patients for clinical trials according to their molecular profile, rather than relying solely on the standard clinical and pathological features adopted to date.