MRI in predicting the response of gastrointestinal stromal tumor to targeted therapy: a patient-based multi-parameter study

Abstract

Background

To investigate the performance of quantitative indicators of MRI in early prediction of the response of gastrointestinal stromal tumor (GIST) to targeted therapy in a patient-based study.

Methods

MRI examinations were performed on 62 patients with GIST using 1.5 T scanners before and at two and 12 weeks after treatment with targeted agents. The longest diameter (LD) and contrast-to-noise ratio (CNR) of the tumors were measured by T2-weighted imaging (T2WI), and the apparent diffusion coefficient (ADC) was determined using diffusion-weighted imaging (DWI). The pre-therapy and early percentage changes (%Δ) of the three parameters were compared with regard to their abilities to differentiate responder and non- responder patients, using ROC curves.

Results

There were 42 patients in responder and 20 in non-responder group. After two weeks of therapy, the percentage changes in the ADC and LD were significantly different between the two groups (ADC: responder 30% vs. non- responder 1%, Z = − 4.819, P < 0.001; LD: responder − 7% vs. non- responder − 2%, Z = − 3.238, P = 0.001), but not in T2WI-CNR (responder − 3% vs. non-responder 9%, Z = − 0.663, P = 0.508). The AUCs on ROC for %ΔLD, %ΔT2WI-CNR and %ΔADC after two weeks of therapy were 0.756, 0.552 and 0.881, respectively, for response differentiation. When %ΔADC ≥15% was used to predict responder, the PPV was 93.3%.

Conclusions

The percentage change of the ADC after two weeks of therapy outperformed T2WI-CNR and longest diameter in predicting the early response of GIST to targeted therapy.

https://ift.tt/2w3gn4V