Implementation of Machine Learning Models to Ensure Radiotherapy Quality for Multicenter Clinical Trials: Report from a Phase III Lung Cancer Study.

The outcome of the patient and the success of clinical trials involving RT is dependent on the quality assurance of the RT plans. Knowledge-based Planning (KBP) models using data from a library of high-quality plans have been utilized in radiotherapy to guide treatment. In this study, we report on the use of these machine learning tools to guide the quality assurance of multicenter clinical trial plans. The data from 130 patients submitted to RTOG1308 were included in this study. Fifty patient cases were used to train separate photon and proton models on a commercially available platform based on principal component analysis. Models evaluated 80 patient cases. Statistical comparisons were made between the KBP plans and the original plans submitted for quality evaluation. Both photon and proton KBP plans demonstrate a statistically significant improvement of quality in terms of organ-at-risk (OAR) sparing. Proton KBP plans, a relatively emerging technique, show more improvements compared with photon plans. The KBP proton model is a useful tool for creating proton plans that adhere to protocol requirements. The KBP tool was also shown to be a useful tool for evaluating the quality of RT plans in the multicenter clinical trial setting.

Optimizing neoadjuvant radiotherapy for resectable and borderline resectable pancreatic cancer using protons.

Approximately 25% of patients diagnosed with pancreatic cancer present with non-metastatic resectable or borderline resectable disease. Unfortunately, the cure rate for these "curable" patients is only in the range of 20%. Local-regional failure rates may exceed 50% after margin-negative, node-negative pancreatectomy, but up to 80% of resections are associated with regional lymph node or margin positivity. While systemic drug therapy and chemotherapy may prevent or delay the appearance of distant metastases, it is unlikely to have a significant impact on local-regional disease control. Preoperative radiotherapy would represent a rational intervention to improve local-regional control. The barrier to preoperative radiotherapy is the concern that it could potentially complicate what is already a long and complicated operation. When the radiotherapy is delivered with X-rays (photons), the entire cylinder of the abdomen is irradiated; therefore, an operating surgeon may be reluctant to accept the associated risk of increased toxicity. When preoperative radiotherapy is delivered with protons, however, significant bowel and gastric tissue-sparing is achieved and clinical outcomes indicate that proton therapy does not increase the risk of operative complications nor extend the length of the procedure.

Radiation-induced tumor immunity in patients with non-small cell lung cancer.

BACKGROUND: Radiation-induced tumor immunity (RITI) influences primary tumor growth and development of metastases in preclinical cancer models with conventional radiotherapy. Antigen-specific immune responses have also been shown for prostate cancer treated with radiotherapy. We examined whether RITI can be induced in patients with non-small cell lung cancer (NSCLC) following proton radiotherapy. METHODS: Pre- and post-radiotherapy plasma samples from 26 patients with nonmetastatic NSCLC who received radiotherapy between 2010 and 2012 were evaluated by western blotting for IgG and IgM bands to assess RITI response to tumor antigens from lung cancer cell lines. Statistical analysis was used to evaluate any correlation among IgG or IgM and clinical outcomes. RESULTS: Twenty-one patients received proton therapy at 2 GyRBE/fraction (n = 17) or 6-12 Gy/fraction (n = 4); five received photon therapy at 2-2.5 GyRBE/fraction. Compared with the pretreatment baseline, new IgG or IgM binding was detected in 27% and 50% of patients, respectively. New IgG bands were detected in the 25-37 kD, 50-75 kD, and 75-100 kD ranges. New IgM bands were detected in the 20-25 kD, 25-37 kD, 37-50 kD, 50-75 kD, and 75-100 kD ranges. There was no difference in IgG and/or IgM RITI response in patients treated with photons versus protons, or in patients who received SBRT compared to standard fractionation (P > 0.05). There was no difference in overall survival, metastasis-free survival, or local control based on IgG and/or IgM RITI response (P > 0.05). CONCLUSION: RITI can be induced in patients with NSCLC through upregulated IgG and/or IgM. RITI response was not associated with proton versus photon therapy or with clinical outcomes in this small cohort and should be examined in a larger cohort in future studies.

Protons offer reduced bone marrow, small bowel, and urinary bladder exposure for patients receiving neoadjuvant radiotherapy for resectable rectal cancer.

BACKGROUND: To assess the potential benefit of proton therapy (PT) over photon therapy, we compared 3-dimensional conformal radiotherapy (3DCRT), intensity-modulated radiotherapy (IMRT), and PT plans in patients undergoing neoadjuvant chemoradiation for resectable rectal cancer at our institution. METHODS: Eight consecutive patients with resectable (T2-T3) rectal cancers underwent 3DCRT, IMRT, and 3-dimensional conformal PT treatment planning. Initial target volumes (PTV1) were contoured using the Radiation Therapy Oncology Group anorectal atlas guidelines. Boost target volumes (PTV2) consisted of the gross rectal tumor plus a uniform 2-cm expansion. Plans delivered 45 Gray (Gy) or Cobalt Gray Equivalent (CGE) to the PTV1 and a 5.4-Gy (CGE) boost to the PTV2. Ninety-five percent of the PTVs received 100% of the target dose and 100% of the PTVs received 95% of the target dose. Standard normal-tissue constraints were utilized. Wilcoxon paired t-tests were performed to compare various dosimetric points between the 3 plans for each patient. RESULTS: All plans met all normal-tissue constraints and were isoeffective in terms of PTV coverage. The proton plans offered significantly reduced median normal-tissue exposure over the 3DCRT and IMRT plans with respect to pelvic bone marrow at the V5Gy, V10Gy, V15Gy, and V20Gy levels and the small bowel space at the V10Gy and V20Gy levels. The proton plans also offered significantly reduced median normal-tissue exposure over the 3DCRT plans with respect to the small bowel at the V30Gy and V40Gy levels and the urinary bladder at the V40Gy level. CONCLUSIONS: By reducing bone marrow exposure, PT may reduce the acute hematologic toxicity of neoadjuvant chemoradiation and increase the likelihood of uninterrupted chemotherapy delivery. Bone marrow sparing may also facilitate the delivery of salvage chemotherapy for patients who subsequently develop hematogenous metastasis. Reduced small bowel exposure using PT may also reduce toxicity and possibly facilitate the use of more-aggressive chemotherapy with radiotherapy.

Androgen deprivation therapy and definitive radiotherapy for prostate cancer.

Adjuvant androgen deprivation therapy (ADT) improves outcomes of patients receiving definitive radiotherapy (RT) for local-regionally advanced prostate cancer. However, patients in most randomized trials had more advanced disease than observed in many practices and were treated with suboptimal RT doses. Although data are conflicting, long-term ADT likely has adverse side-effects in patients with comorbidities. We recommend 6 months of ADT monotherapy with gonadotropin-releasing hormone agonist and RT for patients with high-risk prostate cancer (≥T2c, Gleason Score 8 to 10, and/or prostate-specific antigen ≥20 ng/mL) with minimal or no comorbidities. Adjuvant ADT for unfavorable intermediate-risk patients with a Gleason Score of 4+3=7 is also reasonable.