Correlation of diffusion weighted MR imaging with the prognosis of locally advanced gastric carcinoma to neoadjuvant chemotherapy
Lei Tang1, Ying-Shi Sun1, Zi-Yu Li2, Xiao-Ting Li 1, Fei Shan2, Zi-Ran Li2, and Jia-Fu Ji2

1Radiology, Peking University Cancer Hospital & Institute, Beijing, China, People's Republic of, 2GI surgery, Peking University Cancer Hospital & Institute, Beijing, China, People's Republic of

Synopsis

The percentage changes of ADC after neoadjuvant chemotherapy of gastric carcinoma have correlation with long-term prognosis. The significantly increased ADC after chemotherapy is more prone to signify long-term survival, and has potential to be a surrogate imaging biomarker for the prediction of the prognosis. ADCentire for the whole lesion is better than ADCmin for high signal area in the prognosis prediction.

PURPOSE

To investigate the correlation of the apparent diffusion coefficient (ADC) on diffusion-weighted MRI (DWI) with the prognosis of locally advanced gastric carcinoma to neoadjuvant chemotherapy (NACT).

METHODS

The institutional review board (IRB) approved this prospective study and written informed consents were obtained from all participating patients. Patients with locally advanced gastric carcinoma were enrolled in this study. MRI examinations were performed before and 2 months after NACT (SOX: Oxaliplatin 85 mg/m2 i.v. for day 1, S-1 40~60 mg P.O. for day 1~14, 3 weeks for one period and two period for one cycle). Prior to diffusion-weighted imaging, all patients underwent examinations with routine gastric MR protocols included FSPGR, SSFSE and FRFSE. DWI was performed using a SS-EPI in the transverse plane (TR/TE, 2750 ms/min; matrix size, 128×128; section thickness, 5 mm; intersection gap, 1 mm; field of view, 360-400 mm; NEX, 4; ASSET with an acceleration factor of 2). ADCs of the whole lesion (ADCentire) and high signal area on DWI (ADCmin) were calculated on DWI, and the cancer thickness (THK) on T2-weighted images was measured. All patients were divided into long-term survival group and poor prognosis group, according to the three-year survival status. The pre-therapy baseline values and early percentage changes (%Δ) of the above parameters were compared between groups. %ΔTHK = (THKpost - THKpre) / THKpre × 100%, %ΔADC = (ADCpost - ADCpre) / ADCpre × 100%. Receiver operating characteristics (ROC) curves were employed to compare the performance of the above parameters in the discrimination of different prognosis groups.

RESULTS

A total of 24 patients enrolled in the study. There were 14 patients of long-term survival group and 10 patients of poor prognosis group. No statistical difference in baseline thickness, ADCmin and ADCentire was displayed between long-term survival group and poor prognosis group [Thickness: 16.7 ± 5.9 mm vs. 20.6 ± 8.6mm, P > 0.05; ADCmin: (1.17 ± 0.23) × 10-3 mm2/s vs. (1.23 ± 0.27) × 10-3 mm2/s, P > 0.05; ADCentire: (1.43 ± 0.20) × 10-3 mm2/s vs. (1.50 ± 0.24) × 10-3 mm2/s, P > 0.05]. The %ΔADCmin and %ΔADCentire both had differences between long-term survival group and poor prognosis group (%ΔADCmin: 21% vs. 5%, P =0.06; %ΔADCentire: 23% vs. 1%, P=0.02), but no statistical difference was found on thickness (%ΔTHK: -17% vs. -16%, P=0.79). Through ROC curves, the AUCs for pre-therapy thickness, ADCmin and ADCentire were 0.693, 0.543 and 0.600, respectively; and AUCs for %ΔTHK, %ΔADCmin and %ΔADCentire were 0.532, 0.729 and 0.779, respectively, in the differentiation of prognosis. Using %ΔADC ≥ 15% to predict long-term survival, the positive predictive value (PPV) for %ΔADCmin was 81.8% and %ΔADCentire was 83.3%. Using %ΔADC ≤ 10% to predict poor prognosis, the PPV for %ΔADCmin was 63.6% and %ΔADCentire was 70.0%.

CONCLUSION

The changes of ADC after NACT of gastric carcinoma have correlation with long-term prognosis; the significantly increased ADC is more prone to signify long-term survival. ADCentire is better than ADCmin in the prognosis prediction.

Acknowledgements

This work was supported by National Natural Science Foundation of China (grant no. 81201215, 81371715) and Natural Science Foundation of Beijing (grant no. 7132039)

References

[1] Tang L, Zhang XP, Sun YS, et al. Gastrointestinal stromal tumors treated with imatinib mesylate: apparent diffusion coefficient in the evaluation of therapy response in patients. Radiology, 2011,258:729-738.

[2] Zhang XP, Tang L, Sun YS, et al. Sandwich sign of Borrmann type 4 gastric cancer on diffusion-weighted magnetic resonance imaging. Eur J Radiol, 2012,81:2481-2486.

[3] De Cobelli F, Giganti F, Orsenigo E, et al. Apparent diffusion coefficient modifications in assessing gastro-oesophageal cancer response to neoadjuvant treatment: comparison with tumour regression grade at histology. Eur Radiol, 2013,23:2165-2174.

Figures

ROC curves of pre-therapy ADC and thickness in the prognosis prediction of neoadjuvant chemotherapy for gastric cancer.

ROC curves of %ΔADC and %ΔTHK in the prognosis prediction of neoadjuvant chemotherapy for gastric cancer.

Long-term survival case (57-year old male patient, poor differentiated adenocarcinoma of the EGJ area). The cancer thickness on FRFSE images decreased from 18mm (Fig 3A) to 17mm (Fig 3D) after treatment, with a percentage change of -5.6%. The pre-therapy ADC entire=1.31x10-3 mm2/s, and ADCmin=1.06x10-3 mm2/s (Fig 3C). The post-therapy ADC entire=1.99x10-3 mm2/s, and ADCmin=1.45x10-3 mm2/s (Fig 3F)；with %ΔADCentire = 51.9% , and ADCmin = 36.8%. The patient is still alive after 3 years follow-up.

Poor prognosis case (50-year old male patient, poor differentiated adenocarcinoma of the antrum). The cancer thickness on FRFSE images decreased from 20mm (Fig 4A) to 18mm (Fig 4D), with a percentage change of -10.0%.The pre-therapy ADC entire=1.38x10-3 mm2/s, and ADCmin=1.19x10-3 mm2/s (Fig 4C). The post-therapy ADC entire=1.29x10-3 mm2/s, and ADCmin=1.00x10-3 mm2/s (Fig 4F)；with %ΔADCentire = -6.5% , and ADCmin = -16.0%. The patient was dead after 6 months for diffuse abdominal metastasis.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
2794