Image-Guided High-Dose-Rate (HDR) Boost Localization Using MRI/MR Spectroscopy: A Correlation Study with Biopsy

Purpose: The purpose of this study is to compare the blind interpretations of magnetic resonance imaging (MRI) sequences, diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC), mapping, and magnetic resonance spectroscopy (MRS) of the prostate, in comparison to prostate biopsy to identify a valid dominant intraprostatic lesion (DIL) for dose escalation using high-dose rate brachytherapy. Methods: MRI/MRS were performed on 20 patients with intermediate risk adenocarcinoma of the prostate. T1W, T2W, DWI-ADC, and MRS sequences were performed at 1.5 T with pelvic and endorectal coils. An experienced radiologist rated the presence of cancer in each sextant by using a dichotomic approach, first on MR standard acquisitions (T1W and T2W), then on DWI-ADC mapping, and later on MRS images. Areas under the receiver’s operating characteristic curve were calculated using a sextant as the unit of analysis. The transrectal ultrasonography-guided biopsy results were used as the reference standard. A table summarizing the MRI/MRS findings was made and compared to the corresponding area in the prostate biopsy report. A perfect match was defined to be the presence of cancer in the same sextant of the MRI/MRS exam and the prostate biopsy. Results: The interpretation of the MRI/MRS exams per sextant was compared to the diagnostic biopsy report. MRI readings were compared with the biopsy as a surrogate for the complete pathology specimen of the prostate. A sensitivity (Sn) of 98.6% (95% confidence interval, 92.2% - 99.9%) and specificity (Sp) of 60.8% (46.1% - 74.2%) were found. The positive and negative predictive values (PPV, NPV) were 77.3% (67.1% - 85.5%) and 96.9% (83.8% - 99.9%), respectively. When MRS readings were compared with biopsy, we found a Sn of 96.4% (87.7% - 99.6%) and Sp of 54.8% (38.7% - 70.2%). The PPV and NPV were 74% (62.4% - 83.6%) and 92% (74% - 99%), respectively. DWI-ADC mapping results were also compared with biopsy. We found a Sn and Sp of 93.7% (84.5% - 98.2%) and 82.1% (66.5% - 92.5%), respectively, and a PPV and NPV of 89.4% (79.4% - 95.6%) and 88.9% (73.9% - 96.9%), respectively. Finally, after combining MRI, MRS, and DWI-ADC mapping, compared with biopsy, we obtained a Sn, Sp, PPV, and NPV of 100% (94.8% - 100%), 49% (34.8% - 63.4%), 72.6% (62.5% - 81.3%), and 100% (86.3% - 100%), respectively. Conclusions: The combination of MRI/MRS is a sensitive tool for both the structural and metabolic evaluation of prostate cancer location. MRI/MRS exams are useful to delineate a DIL for high-dose-rate (HDR) intraprostatic boost.