We performed a retrospective analysis out of a GISTs pathological database from our institution,
focusing on the last year,
selecting the most iconic cases for educational purposes.
Contrast-enhanced abdominal and pelvic computed tomography (CT) is the preferred imaging technique in what concerns disease staging,
follow-up evaluations and assessment of TKI responses.4,9 CT is also the modality of choice to evaluate putative complications related to surgery and to assist core biopsies that cannot be performed through endoscopy.
Since GISTs are hypervascular lesions,
the adequate CT protocol should include negative oral contrast with the acquisition of a triphasic CT (non-enhanced,
arterial phase and portovenous phase).
The arterial phase is crucial in order to evaluate hepatic involvement,
since some lesions may not be visible in the portovenous phase.9
18FDG PET-TC should be performed when CT findings are not conclusive or not coherent with clinical presentation and eventually in the evaluation of disseminated disease,
once it is considered a research tool,
thus not contributing to GISTs staging.9,10
Staging
Both tumour size and site of origin are two out of the three well established independent prognostic factors,
as well as mitotic rate,
and imaging is a significant part of staging workup.4 Radiologic presentations of GISTs can differ based on the tumour size.
Since GISTs arise from muscularis propria,
they commonly show an exophytic or intramural growth pattern (Fig. 2).
Exophytic lesions can invade adjacent structures,
such as the pancreas,
the colon or the diaphragm.2 Thus,
the imaging main purpose is to evaluate the extent of the primary tumour and to rule out intra-abdominal metastases.10
Most GISTs have a spherical or ovoid morphology,
and its maximum diameter should be determined.
The thresholds of the greatest tumour diameter used in TNM classification are 2,
5 and 10cm.
They are frequently large,
heterogeneous and hypervascular (enhancing) lesions which may present areas of haemorrhage,
necrosis or cystic degeneration,
appearing as low attenuation areas (Fig. 3).
These areas of necrosis and/or haemorrhage may cavitate,
creating a fistula to gastrointestinal tract.
Although rare,
calcified clusters can also be seen (Fig. 4).2 GISTs are usually well circumscribed lesions with lobulated contours due to the absence of surrounding desmoplastic reaction.
However,
small lesions (< 2 cm) are usually considered incidentalomas and “benign”.
They usually present as round and well-defined lesions,
with homogeneous enhancement,
which may introduce other mesenchymal tumours as differential diagnosis (such as leiomyomas,
neuroendocrine tumours or schwannomas) (Fig. 5).2,4
The origin in the stomach is a positive prognostic factor,
being more frequent in its body.
As mentioned before,
other known locations are the small intestine jejunum and the ileum (30%),
the duodenum (5%) (Fig. 6),
the rectum (3%) the colon (1%),
the oesophagus (<1%),
the omentum and the mesentery.
Nodal involvement is rare,
except for succinate dehydrogenase-deficient (SDH) GISTs (most of paediatric GISTs).9,10
Approximately 50% of patients with GISTs present with metastatic involvement,
most frequently involving the liver and the peritoneum through haematogenous spread and peritoneal seeding (Fig. 7),
respectively.11 Metastases are similar to primary GISTs with respect to CT characteristics and tumours vessels are frequently seen within the tumour (Fig. 8).9
Given this,
initial CT should give fundamental information related to GIST staging (Table 1). Risk stratification is based on NCCN Task Force report and depends on tumour site,
size and mitotic index.12
Response evaluation
Surgical resection is the definitive treatment for primary GISTs,
but its recurrence occurs in most patients,
independently of an eventual tumour-free margin.
After a resection or a TKI inhibitor response,
imaging plays an important role as a surveillance tool to detect recurrence or disease progression.9
Response assessment (to treatment with TKI inhibitors) if performed through CT and the best parameters which correlate with therapy response are not only the tumour site but also its attenuation,
since some tumours may demonstrate density changes that precede tumour shrinkage.
In few cases the tumour may even enlarge due to haemorrhage or mixoid degeneration (“false progression”).
The classical volumetric response takes some months before fitting Response Evaluation Criteria in Solid Tumours (RECIST) and is typically characterized by a transition from a heterogeneously hyperattenuating pattern into a homogeneously hypoattenuating one,
with less tumours vessels and resolution of tumour “nodules” (Fig. 9 and Fig. 10).9 Therefore,
the response criteria proposed by Choi et al.
are more adequate than RECIST,
since they take the tumour density change into account,
whereas RECIST does not.13 8FDG PET-TC is very sensitive for early assessment of response and may be performed in few cases,
such as cytoreductive treatments.4
It is important to bear in mind that an enhancing nodule within a treated hypoattenuated tumour indicates GIST recurrence (“nodule within the mass” sign),
which is a unique feature of this condition.
Other classical indicators of GIST recurrence/progression are an increase in the tumour size,
an increase in the overall density of the lesion,
the presence of new lesions at the site of a primary disease and the development of metastases.9 There are no optimal imaging follow-up intervals,
but ESMO–EURACAN Clinical Practice Guidelines give example of some institutions (Table 2).