A multi-centric evaluation of self-learning GAN based pseudo-CT generation software for low field pelvic magnetic resonance imaging.

Purpose/objectives: An artificial intelligence-based pseudo-CT from low-field MR images is proposed and clinically evaluated to unlock the full potential of MRI-guided adaptive radiotherapy for pelvic cancer care. Materials and method: In collaboration with TheraPanacea (TheraPanacea, Paris, France) a pseudo-CT AI-model was generated using end-to-end ensembled self-supervised GANs endowed with cycle consistency using data from 350 pairs of weakly aligned data of pelvis planning CTs and TrueFisp-(0.35T)MRIs. The image accuracy of the generated pCT were evaluated using a retrospective cohort involving 20 test cases coming from eight different institutions (US: 2, EU: 5, AS: 1) and different CT vendors. Reconstruction performance was assessed using the organs at risk used for treatment. Concerning the dosimetric evaluation, twenty-nine prostate cancer patients treated on the low field MR-Linac (ViewRay) at Montpellier Cancer Institute were selected. Planning CTs were non-rigidly registered to the MRIs for each patient. Treatment plans were optimized on the planning CT with a clinical TPS fulfilling all clinical criteria and recalculated on the warped CT (wCT) and the pCT. Three different algorithms were used: AAA, AcurosXB and MonteCarlo. Dose distributions were compared using the global gamma passing rates and dose metrics. Results: The observed average scaled (between maximum and minimum HU values of the CT) difference between the pCT and the planning CT was 33.20 with significant discrepancies across organs. Femoral heads were the most reliably reconstructed (4.51 and 4.77) while anal canal and rectum were the less precise ones (63.08 and 53.13). Mean gamma passing rates for 1%1mm, 2%/2mm, and 3%/3mm tolerance criteria and 10% threshold were greater than 96%, 99% and 99%, respectively, regardless the algorithm used. Dose metrics analysis showed a good agreement between the pCT and the wCT. The mean relative difference were within 1% for the target volumes (CTV and PTV) and 2% for the OARs. Conclusion: This study demonstrated the feasibility of generating clinically acceptable an artificial intelligence-based pseudo CT for low field MR in pelvis with consistent image accuracy and dosimetric results.

Magnetic Resonance Guided Radiation Therapy for Pancreatic Adenocarcinoma, Advantages, Challenges, Current Approaches, and Future Directions.

INTRODUCTION: Pancreatic adenocarcinoma (PAC) has some of the worst treatment outcomes for any solid tumor. PAC creates substantial difficulty for effective treatment with traditional RT delivery strategies primarily secondary to its location and limited visualization using CT. Several of these challenges are uniquely addressed with MR-guided RT. We sought to summarize and place into context the currently available literature on MR-guided RT specifically for PAC. METHODS: A literature search was conducted to identify manuscript publications since September 2014 that specifically used MR-guided RT for the treatment of PAC. Clinical outcomes of these series are summarized, discussed, and placed into the context of the existing pancreatic literature. Multiple international experts were involved to optimally contextualize these publications. RESULTS: Over 300 manuscripts were reviewed. A total of 6 clinical outcomes publications were identified that have treated patients with PAC using MR guidance. Successes, challenges, and future directions for this technology are evident in these publications. MR-guided RT holds theoretical promise for the treatment of patients with PAC. As with any new technology, immediate or dramatic clinical improvements associated with its use will take time and experience. There remain no prospective trials, currently publications are limited to small retrospective experiences. The current level of evidence for MR guidance in PAC is low and requires significant expansion. Future directions and ongoing studies that are currently open and accruing are identified and reviewed. CONCLUSIONS: The potential promise of MR-guided RT for PAC is highlighted, the challenges associated with this novel therapeutic intervention are also reviewed. Outcomes are very early, and will require continued and long term follow up. MR-guided RT should not be viewed in the same fashion as a novel chemotherapeutic agent for which dosing, administration, and toxicity has been established in earlier phase studies. Instead, it should be viewed as a novel procedural intervention which must be robustly tested, refined and practiced before definitive conclusions on the potential benefits or detriments can be determined. The future of MR-guided RT for PAC is highly promising and the potential implications on PAC are substantial.

Changes in neutrophil-to-lymphocyte and platelet-to-lymphocyte ratios during chemoradiation predict for survival and pathologic complete response in trimodality esophageal cancer patients.

BACKGROUND: Neutrophil-to-lymphocyte ratio (NLR) and platelet-to-lymphocyte ratio (PLR) predict for survival in cancer patients. In patients receiving multimodality therapy, the effect of therapy on the NLR and PLR is not well understood. We evaluated changes in NLR and PLR among locally advanced esophageal cancer patients who received trimodality therapy. METHODS: We performed a retrospective analysis of nonmetastatic patients with esophageal cancer who received neoadjuvant chemoradiation therapy (CRT) followed by esophagectomy at our institution between March 2000 and April 2012. NLR and PLR values were obtained for the following time points (TPs): (I) at diagnosis before CRT; (II) after CRT but prior to surgery; and (III) after surgery. We evaluated changes in NLR and PLR using the difference and ratio between TPs. Overall survival (OS) was evaluated by Kaplan-Meier analysis. Univariate and multivariate Cox regression models were applied to evaluate the independent prognostic significance of NLR and PLR. RESULTS: This IRB-approved study included the records of 83 consecutive patients with stage II-IV esophageal cancer. The median age was 60 years, and median follow-up was 29.3 months. Patients were treated to a median prescription dose of 50.4 Gy (range, 50.4-56.4 Gy) in 28-33 fractions. Median NLR and PLR were 3.3 and 157.2, 12 and 645, and 11.5 and 391.7 at TPs 1, 2, and 3, respectively. On multivariate analysis, superior OS was associated with PLR ≥250 at TP3 (P=0.03), PLR decrease ≥609.2 between TP2 and TP3 (P=0.02), and PLR ratio (TP3/TP1) ≥1.08 (P=0.03). Inferior progression-free survival (PFS) was associated with NLR ≥36 at TP2 (P=0.0008), NLR increase ≥28.3 between TP1 and TP2 (P=0.0005), and PLR ratio (TP2/TP3) ≥0.38 (P=0.1). Pathologic complete response (PCR) was less likely for adenocarcinoma (AC) histology (P=0.03), NLR ≥10.6 at TP2 (P=0.04), and NLR increase ≥4.6 from TP1 to TP2 (P=0.03). CONCLUSIONS: To our knowledge, this is the first study to examine NLR and PLR values at various time intervals throughout treatment and demonstrate a correlation between OS, PFS, and PCR in patients undergoing trimodality therapy for esophageal cancer.