Adam Gladwish MD, MSc, Presenter: Nothing to Disclose

Michael Milosevic, Abstract Co-Author: Nothing to Disclose

Anthony Fyles, Abstract Co-Author: Nothing to Disclose

Warren D. Foltz PhD, Abstract Co-Author: Nothing to Disclose

Nathan Becker PhD, Abstract Co-Author: Nothing to Disclose

Haiyan Jiang, Abstract Co-Author: Nothing to Disclose

Wilfred Levin MD, Abstract Co-Author: Nothing to Disclose

Lee Manchul MD, Abstract Co-Author: Nothing to Disclose

Kathy Han MD, Abstract Co-Author: Nothing to Disclose

To investigate whether pre-treatment apparent diffusion coefficient (ADC) from diffusion-weighted MRI is predictive of disease recurrence in women with locally advanced cervical cancer treated with radical chemoradiation.

Eighty-five women with stage IB-IVA cervical cancer treated with radical chemoradiation from 2009-2013 had staging MRI, including T2-weighted (T2W) and DWI series at either 1.5 or 3.0T (b-value 0-1000). T2W images and ADC maps were co-registered in Pinnacle, and the tumor was delineated on the ADC maps with the aid of T2W images. The mean, median, 75th, 90th and 95th percentile ADC of the tumor were extracted using Matlab, and normalized to each patient's mean urine ADC to reduce variability (reported as dimensionless quantitates: nADCmean, nADCmedian, nADC75, nADC90, nADC95, respectively). The primary outcome was disease-free survival (DFS). Information about recurrence and survival was recorded prospectively. Uni- and multivariate Cox regression analyses evaluated the predictive value of ADC parameters and relevant clinical variables (age, histology, stage, nodal status, baseline hemoglobin, tumor size) for DFS.

Of the 85 women included, 62 were disease free at last follow-up (FU). Median FU was 2.2 years. FIGO stage was IB in 45 patients, II in 33, and III/IVA in 7 patients. Median tumor diameter was 4.8 cm. Thirty-two women had pelvic nodal involvement, and 11 had paraaortic nodal involvement on staging CT/MR. The median nADC95 was 0.58 (range 0.41-0.83). Significant variables on univariate analysis included tumor size (HR 1.45, p = 0.01), paraaortic nodal involvement (HR 4.25, p=0.002), and all nADC parameters (HR ranged from 0.91-0.94 per 0.01 increase in nADC, p < 0.04). nADC75, nADC90, and nADC95 were highly correlated, and therefore only nADC95 (lowest HR and p value) was included in multivariate analysis. On multivariate analysis, paraaortic nodal involvement and nADC95 remained predictive of DFS (HR 3.12, p=0.02; and HR 0.92, p=0.005, respectively).

ADC may be a useful imaging biomarker for predicting treatment failure in patients with locally advanced cervical cancer treated with chemoradiation.

Identification of patients with locally advanced cervical cancer at higher risk of recurrence based on the pre-treatment apparent diffusion coefficient will enable personalized treatment approach.

Gladwish, A, Milosevic, M, Fyles, A, Foltz, W, Becker, N, Jiang, H, Levin, W, Manchul, L, Han, K, Does the Apparent Diffusion Coefficient Value Predict Disease Recurrence in Patients with Locally Advanced Cervical Cancer Treated with Radical Chemoradiation?.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14010872.html