Shaped-Charge Learning Architecture for the Human-Machine Teams.

In spite of great progress in recent years, deep learning (DNN) and transformers have strong limitations for supporting human-machine teams due to a lack of explainability, information on what exactly was generalized, and machinery to be integrated with various reasoning techniques, and weak defense against possible adversarial attacks of opponent team members. Due to these shortcomings, stand-alone DNNs have limited support for human-machine teams. We propose a Meta-learning/DNN → kNN architecture that overcomes these limitations by integrating deep learning with explainable nearest neighbor learning (kNN) to form the object level, having a deductive reasoning-based meta-level control learning process, and performing validation and correction of predictions in a way that is more interpretable by peer team members. We address our proposal from structural and maximum entropy production perspectives.

Low-dose preoperative radiation, resection, and reduced-field postoperative radiation for soft tissue sarcomas.

BACKGROUND AND OBJECTIVES: Radiotherapy (RT) enables conservative surgery for soft tissue sarcoma (STS). RT can be delivered either pre-operatively (PreRT) or postoperatively (PORT), yet in some patients, neither approach is fully satisfactory (e.g., urgent surgery or wound healing risk prevents PreRT, yet PORT alone cannot cover the entire surgical field). We hypothesized that, in such situations, low-dose PreRT (LD-PreRT) would decrease the risk of intraoperative tumor seeding and thus permit PORT to a reduced volume (covering the high-risk tumor bed but not all surgically manipulated tissues). METHODS: We identified a single-institution retrospective cohort of 78 patients treated with LD-PreRT (10-30 Gy), resection, and PORT between 1980 and 2018. RESULTS: At a median follow-up of 8.2 years, 8-year overall survival (OS) was 65.9%, disease-free survival (DFS) 50.5%, and local control (LC) 76.7%; in 45 patients with extremity/superficial trunk (E/ST) STS, 8-year LC was 80.9%. Both before and after propensity score adjustment, there were no differences in OS, DFS, or LC between this cohort and a separate cohort of 394 STS (221 E/ST-STS) patients treated with surgery and PORT alone. CONCLUSIONS: In patients for whom neither PreRT nor PORT alone is optimal, LD-PreRT may prevent intraoperative tumor seeding and enable PORT to a reduced volume while preserving oncologic outcomes.

Quality of life in patients with proton-treated pediatric medulloblastoma: Results of a prospective assessment with 5-year follow-up.

BACKGROUND: To the authors' knowledge, health-related quality of life (HRQOL) outcomes are not well described in patients with medulloblastoma. The use of proton radiotherapy (RT) may translate into an improved HRQOL. In the current study, the authors report long-term HRQOL in patients with proton-treated pediatric medulloblastoma. METHODS: The current study was a prospective cohort HRQOL study of patients with medulloblastoma who were treated with proton RT and enrolled between August 5, 2002, and October 8, 2015. Both child report and parent-proxy report Pediatric Quality of Life Inventory (PedsQL) surveys were collected at baseline during RT and annually thereafter (score range on surveys of 0-100, with higher scores indicating better HRQOL). Patients were dichotomized by clinical/treatment variables and subgroups were compared. Mixed-model analysis was performed to determine the longitudinal trajectory of PedsQL scores. The Student t test was used to compare long-term HRQOL measures with published means from a healthy child population. RESULTS: Survey data were evaluable for 116 patients with a median follow-up of 5 years (range, 1-10.6 years); the median age at the time of diagnosis was 7.6 years (range, 2.1-18.1 years). At baseline, children reported a total core score (TCS) of 65.9, which increased by 1.8 points annually (P<.001); parents reported a TCS of 59.1, which increased by 2.0 points annually. Posterior fossa syndrome adversely affected baseline scores, but these scores significantly improved with time. At the time of last follow-up, children reported a TCS of 76.3, which was 3.3 points lower than that of healthy children (P = .09); parents reported a TCS of 69, which was 11.9 points lower than that of parents of healthy children (P<.001). Increased follow-up time from diagnosis correlated with improved HRQOL scores. CONCLUSIONS: HRQOL scores appear to increase over time after treatment in children treated with proton RT for medulloblastoma but remain lower compared with those of parent-proxy reports as well as published means from a healthy normative sample of children. Additional follow-up may translate into continued improvements in HRQOL. Cancer 2018. © 2018 American Cancer Society.

Long-term outcomes among breast cancer patients with extensive regional lymph node involvement: implications for locoregional management.

Extensive lymph node (LN) involvement portends significant risk for distant metastasis (DM) among breast cancer patients. As a result, local management may be of secondary import to systemic control in this population. We analyzed patients with ≥10 involved LNs (N3) to evaluate the feasibility of breast conserving therapy (BCT) vs modified radical mastectomy (MRM) in this high-risk cohort. Among 98 women with N3 disease 46 (46.9%) underwent BCT and 52 (53.1%) received MRM. Nearly all patients (92%) received comprehensive radiotherapy (RT) including axillary and supraclavicular fields. The Kaplan-Meier method and Cox regression analyses were used to analyze time-to-event outcomes. Median follow-up was 76 months, with a 5-year DFS of 64.9% and OS of 71.9% among the cohort. Poorly differentiated (p = 0.007), ER-negative tumors (p = 0.015) had adverse DFS outcomes. Treatment groups did not differ with regard to 10-year DFS (45.4% for MRM vs. 57.6% for BCT; p = 0.31), or OS (61.4 vs. 63.7%; p = 0.79). DM-free survival was 48.9% following MRM and 60.6% following BCT (p = 0.19). Patients with ≥10 involved LNs have similar outcomes following BCT or MRM, suggesting that RT may obviate the need for more-extensive surgery. While local control is comparably favorable regardless of surgical approach, systemic control remains a challenge in this population.

MRI surveillance following treatment of extremity soft tissue sarcoma.

BACKGROUND AND OBJECTIVES: Local recurrence (LR) following limb-sparing surgery and radiation therapy (RT) for extremity soft tissue sarcoma (STS) is rare. The current study investigates the utility of surveillance nuclear magnetic resonance imaging (MRI) for detection of asymptomatic LRs. METHODS: The study cohort consisted of 168 adult patients with extremity STS treated with limb-sparing surgery and RT with curative intent between October 2001 and January 2011. Follow-up surveillance MRIs and history and physical examinations were performed per the NCCN guidelines with additional MRIs as clinically indicated. The method of LR detection and MRI number and indication were determined. RESULTS: After a median follow-up of 4.7 years (range: 0.6-10.5) 11 (6.5%; 11/168) patients developed LRs. Five hundred two MRIs were obtained, 429 (85.5%; 429/502) for surveillance and 73 (14.5%; 73/502) as clinically indicated. One hundred fourteen patients underwent ≥1 surveillance MRI. The median surveillance MRI interval was 6.4 months (range 1.4-68.9). Surveillance MRI detected an asymptomatic LR in 1 (0.9%; 1/114) patient with a complex reconstruction. CONCLUSIONS: Surveillance MRI infrequently detects asymptomatic LRs following limb-sparing surgery and RT for extremity STS and should be limited to patients whose primary tumor sites are not easily assessed by history and physical examination.

Role of post-operative radiation boost for soft tissue sarcomas with positive margins following pre-operative radiation and surgery.

BACKGROUND AND OBJECTIVES: The role of a radiation therapy (RT) boost for positive margins following pre-operative RT and surgery in extremity soft tissue sarcomas (STS) is unclear. We assessed the contribution of a boost to local control (LC), disease-free survival (DFS), and overall survival (OS). METHODS: We identified 67 patients treated from 1987 to 2011 with pre-operative RT and surgery with positive margin(s). Select patients received a boost delivered as peri-operative Iridium-192 brachytherapy (BRT), intra-operative electrons (IORT), or post-operative external beam RT (EBRT). RESULTS: Ten patients received no RT boost, 10 received a BRT or IORT boost, and 47 received an EBRT boost. Five-year LC rates for no boost, BRT/IORT boost, and EBRT boost were 100%, 78%, and 71% (P = 0.5). On multivariate analysis, there were no significant predictors for LC. Variables associated with improved DFS rates were single positive margin (P = 0.007) and low tumor grade (P = 0.03). Tumor size <5 cm (P = 0.003), low grade (P = 0.001), and boost (P = 0.02) were associated with longer survival. CONCLUSIONS: We did not identify a LC advantage for an RT boost. Given the unidentified selection factors for delivery of boost and its potential toxicities, its role in this setting remains unproven.

Disparities in Penile Cancer Incidence, Mortality, and Place of Death Trends From 1999 to 2020.

INTRODUCTION: Penile cancer is rare in the United States (US); however, disparities have been found in the incidence, treatment, and outcomes of penile cancer. There is a need for evaluation of recent trends in penile cancer mortality, incidence, and place of death across all demographics. MATERIALS AND METHODS: Using the CDC WONDER database, penile cancer-specific mortality (PNCSM) trends in the US were evaluated from 1999 to 2020 by race/ethnicity, age group, census region, and place of death. Penile cancer incidence trends for the US from 1995 to 2019 were gathered from the NAACCR database. Average annual percent changes for mortality and incidence rates were determined using Joinpoint regression modeling. Univariable and multivariable logistic regression were used to evaluate independent predictors associated with place of death. RESULTS: From 1999 to 2020, 5833 people died from penile cancer in the US. Overall PNCSM increased by 1.8% per year from 1999-2020 (95% CI, 1.3%, 2.2%). Non-Hispanic White patients and Hispanic patients had increasing PNCSM rates from 1999-2020 (2.1 [95% CI, 1.5%, 2.7%]; 1.9 [95% CI, 1.0%, 2.8%], respectively). From the place of death analysis, Hispanic patients were at higher odds of dying at home or hospice when compared to non-Hispanic White patients (adjusted odds ratio [aOR] = 1.19, P = .045). Age-adjusted incidence rates for all stages of penile cancer increased significantly from 1995-2016 (AAPC, 0.7% [95% CI, 0.4%, 1.0%]), driven by regional and distant penile cancer incidence rates (AAPC 1995-2019, regional: 2.0% [95% CI, 1.7%, 2.4%]; AAPC 1995-2019, distant: 2.5% [95% CI, 1.8%, 3.1%]). CONCLUSION: The increasing penile cancer-specific mortality and incidence rates indicate the need for further improvements in screening, diagnosis, and treatment. Widespread efforts across all demographics are needed to ensure early detection of the disease.

The radiopharmaceutical radium-223 has immunomodulatory effects in patients and facilitates anti-programmed death receptor-1 therapy in murine models of bone metastatic prostate cancer.

BACKGROUND & PURPOSE: Radium-223 (Ra223) improves survival in metastatic prostate cancer (mPC), but its impact on systemic immunity is unclear, and biomarkers of response are lacking. We examined markers of immunomodulatory activity during standard clinical Ra223 and studied the impact of Ra223 on response to immune checkpoint inhibition (ICI) in preclinical models. MATERIALS & METHODS: We conducted a single-arm biomarker study of Ra223 in 22 bone mPC patients. We measured circulating immune cell subsets and a panel of cytokines before and during Ra223 therapy and correlated them with overall survival (OS). Using two murine mPC models-orthotopic PtenSmad4-null and TRAMP-C1 grafts in syngeneic immunocompetent mice-we tested the efficacy of combining Ra223 with ICI. RESULTS: Above-median level of IL-6 at baseline was associated with a median OS of 358 versus 947 days for below levels; p = 0.044, from the log-rank test. Baseline PlGF and PSA inversely correlated with OS (p = 0.018 and p = 0.037, respectively, from the Cox model). Ra223 treatment was associated with a mild decrease in some peripheral immune cell populations and a shift in the proportion of MDSCs from granulocytic to myeloid. In mice, Ra223 increased the proliferation of CD8+ and CD4+ helper T cells without leading to CD8+ T cell exhaustion in the mPC lesions. In one of the models, combining Ra223 and anti-PD-1 antibody significantly prolonged survival, which correlated with increased CD8+ T cell infiltration in tumor tissue. CONCLUSION: The inflammatory cytokine IL-6 and the angiogenic biomarker PlGF at baseline were promising outcome biomarkers after standard Ra223 treatment. In mouse models, Ra223 increased intratumoral CD8+ T cell infiltration and proliferation and could improve OS when combined with anti-PD-1 ICI.

Proton-Based Radiation Therapy for Skull Base Chondrosarcomas in Children and Adolescents: 40-year experience from a single institution.

BACKGROUND: The literature on skull base chondrosarcoma (CHS) is scarce. We report outcomes for pediatric and young patients ≤ 22 years with base of skull (BOS) CHSs treated with proton-based radiotherapy (PBRT). METHODS: We retrospectively reviewed all patients treated with PBRT between 1981 and 2023. Primary endpoints were overall survival (OS), chondrosarcoma-specific survival (CSS), progression-free survival (PFS), local control (LC), and distal control (DC). RESULTS: Eighty-four patients were identified. Median age at diagnosis was 19 years (range, 6 - 22). Most patients (n=79, 94%) had conventional CHS while 5 (6%) had mesenchymal CHS. Nine patients (11%) underwent gross total resection (GTR), 64 (76%) subtotal resection (STR) and 11 patients (13%) underwent biopsy. Twenty-four (29%) patients progressed prior to radiotherapy (RT). The median prescription dose was 70 Gy (RBE) (range, 50 - 79.7). At a median follow-up of 18.0 years [interquartile range (IQR),), 9.2 -26.2] from diagnosis, 11 patients recurred (7 local, 4 distant). Six patients died of disease. Five patients died of other causes. Ten-year OS, CSS and PFS rates for all patients were 93.3%, 94.7% and 88%, respectively. Twenty-year OS, CSS and PFS rates for conventional CHS (n=79) were 93.1%, 97.1% and 89.2%, respectively. Mesenchymal histology was significantly associated with worse OS and PFS. Pre-RT tumor progression portended worse OS and CSS. Eight patients (9.9 %) experienced a late toxicity grade 3 or greater. CONCLUSIONS: This is the largest cohort of pediatric BOS CHSs in literature to date. High-dose PBRT following surgical resection achieves excellent disease control with minimal toxicity.

Updated long-term outcomes and prognostic factors for patients with unresectable locally advanced pancreatic cancer treated with intraoperative radiotherapy at the Massachusetts General Hospital, 1978 to 2010.

BACKGROUND: In the current study, the authors evaluated long-term outcomes, intraoperative radiotherapy (IORT)-related toxicity, and prognostic factors for overall survival (OS) among patients with unresectable locally advanced pancreatic cancer (LAPC) who received IORT as part of their treatment at the Massachusetts General Hospital (MGH). METHODS: Medical records were reviewed for 194 consecutive patients with unresectable LAPC who were treated with IORT at MGH between 1978 and 2010. OS was calculated using the Kaplan-Meier method. Prognostic factors were evaluated at the univariate level by the log-rank test and at the multivariate level by the Cox proportional hazards model. Rates of disease progression and treatment toxicity were calculated. RESULTS: The 1-year, 2-year, and 3-year survival rates were 49%, 16%, and 6%, respectively. Six patients (3%) survived for > 5 years. The median OS was 12.0 months. Among 183 patients with known post-IORT disease status, the 2-year local progression-free survival and distant metastasis-free survival rates were 41% and 28%, respectively. On multivariate analysis, an IORT applicator diameter ≤ 8 cm (hazards ratio [HR], 0.51; 95% confidence interval [95% CI], 0.30-0.84 [P = .009]), a Charlson age-comorbidity index ≤ 3 (HR, 0.47; 95% CI, 0.31-0.73 [P = .001]), and receipt of chemotherapy (HR, 0.46; 95% CI, 0.33-0.66 [P < .001]) predicted improved OS. The median OS for patients with all 3 positive prognostic factors was 21.2 months. CONCLUSIONS: Well-selected patients with LAPC with small tumors and low Charlson age-comorbidity indices can achieve good long-term survival outcomes with a treatment regimen that incorporates chemotherapy and IORT.

Using machine learning for healthcare treatment planning.

We present a methodology for using machine learning for planning treatments. As a case study, we apply the proposed methodology to Breast Cancer. Most of the application of Machine Learning to breast cancer has been on diagnosis and early detection. By contrast, our paper focuses on applying Machine Learning to suggest treatment plans for patients with different disease severity. While the need for surgery and even its type is often obvious to a patient, the need for chemotherapy and radiation therapy is not as obvious to the patient. With this in mind, the following treatment plans were considered in this study: chemotherapy, radiation, chemotherapy with radiation, and none of these options (only surgery). We use real data from more than 10,000 patients over 6 years that includes detailed cancer information, treatment plans, and survival statistics. Using this data set, we construct Machine Learning classifiers to suggest treatment plans. Our emphasis in this effort is not only on suggesting the treatment plan but on explaining and defending a particular treatment choice to the patient.

The RNA-Seq data analysis shows how the ontogenesis defines aging.

This paper presents a global statistical analysis of the RNA-Seq results of the entire Mus musculus genome. We explain aging by a gradual redistribution of limited resources between two major tasks of the organism: its self-sustenance based on the function of the housekeeping gene group (HG) and functional differentiation provided by the integrative gene group (IntG). All known disorders associated with aging are the result of a deficiency in the repair processes provided by the cellular infrastructure. Understanding exactly how this deficiency arises is our primary goal. Analysis of RNA production data of 35,630 genes, from which 5,101 were identified as HG genes, showed that RNA production levels in the HG and IntG genes had statistically significant differences (p-value <0.0001) throughout the entire observation period. In the reproductive period of life, which has the lowest actual mortality risk for Mus musculus, changes in the age dynamics of RNA production occur. The statistically significant dynamics of the decrease of RNA production in the HG group in contrast to the IntG group was determined (p-value = 0.0045). The trend toward significant shift in the HG/IntG ratio occurs after the end of the reproductive period, coinciding with the beginning of the mortality rate increase in Mus musculus indirectly supports our hypothesis. The results demonstrate a different orientation of the impact of ontogenesis regulatory mechanisms on the groups of genes representing cell infrastructures and their organismal functions, making the chosen direction promising for further research and understanding the mechanisms of aging.

Disparities in testicular cancer incidence, mortality, and place of death trends from 1999 to 2020: A comprehensive cohort study.

BACKGROUND: Testicular cancer (TC) mortality rates have decreased over time, however it is unclear whether these improvements are consistent across all communities. AIMS: The aim of this study was to analyze trends in TC incidence, mortality, and place of death (PoD) in the United States between 1999-2020 and identify disparities across race, ethnicity, and geographic location. METHODS AND RESULTS: This cross-sectional study used CDC WONDER and NAACCR, to calculate age-adjusted rates of TC incidence and mortality, respectively. PoD data for individuals who died of TC were collected from CDC WONDER. Using Joinpoint analysis, longitudinal mortality trends were evaluated by age, race, ethnicity, US census region, and urbanization category. TC stage (localized vs metastatic) trends were also evaluated. Univariate and multivariate regression analysis identified demographic disparities for PoD. A total of 8,456 patients died of TC from 1999-2020. Average annual percent change (AAPC) of testicular cancer-specific mortality (TCSM) remained largely stable (AAPC, 0.4; 95% CI -0.2 to 0.9; p = 0.215). Men ages 25-29 experienced a significant increase in TCSM (AAPC, 1.3, p = 0.003), consistent with increased metastatic testicular cancer-specific incidence (TCSI) trend for this age group (AAPC, 1.6; p < 0.01). Mortality increased for Hispanic men (AAPC, 1.7, p < 0.001), with increased metastatic TCSI (AAPC, 2.5; p < 0.001). Finally, younger (<45), single, and Hispanic or Black men were more likely to die in medical facilities (all p < 0.001). The retrospective study design is a limitation. CONCLUSION: Significant increases in metastatic TC were found for Hispanic men and men aged 25-29 potentially driving increasing testicular cancer specific mortality in these groups. Evidence of racial and ethnic differences in place of death may also highlight treatment disparities.

US Cancer Mortality Trends Among Hispanic Populations From 1999 to 2020.

Importance: Advances in cancer research and treatment access have led to decreasing cancer mortality in the US; however, cancer remains the leading cause of death among Hispanic individuals. Objective: To evaluate longitudinal cancer mortality trends from 1999 to 2020 among Hispanic individuals by demographic characteristics and to compare age-adjusted cancer death rates between the Hispanic population and other racial and ethnic populations during 2000, 2010, and 2020. Design, Setting, and Participants: This cross-sectional study obtained age-adjusted cancer death rates among Hispanic individuals of all ages between January 1999 and December 2020, using the Centers for Disease Control and Prevention WONDER database. Cancer death rates in other racial and ethnic populations were extracted for 2000, 2010, and 2020. Data were analyzed from October 2021 to December 2022. Exposures: Age, gender, race, ethnicity, cancer type, and US census region. Main Outcomes and Measures: Trends and average annual percent changes (AAPCs) in age-adjusted cancer-specific mortality (CSM) rates among Hispanic individuals were estimated by cancer type, age, gender, and region. Results: From 1999 to 2020, 12 644 869 patients died of cancer in the US, of whom 690 677 (5.5%) were Hispanic; 58 783 (0.5%) were non-Hispanic American Indian or Alaska Native; 305 386 (2.4%), non-Hispanic Asian or Pacific Islander; 1 439 259 (11.4%), non-Hispanic Black or African American; and 10 124 361 (80.1%), non-Hispanic White. For 26 403 patients (0.2%), no ethnicity was stated. The overall CSM rate among Hispanic individuals decreased by 1.3% (95% CI, 1.2%-1.3%) annually. Overall CSM rate decreased more for Hispanic men (AAPC, -1.6%; 95% CI, -1.7% to -1.5%) compared with women (AAPC, -1.0%; 95% CI, -1.0% to -0.9%). While death rates among Hispanic individuals decreased for most cancer types, mortality rates for liver cancer (AAPC, 1.0%; 95% CI, 0.6%-1.4%) increased among Hispanic men, and rates of liver (AAPC, 1.0%; 95% CI, 0.8%-1.3%), pancreas (AAPC, 0.2%; 95% CI, 0.1%-0.4%), and uterine (AAPC, 1.6%; 95% CI, 1.0%-2.3%) cancers increased among Hispanic women. Overall CSM rates increased for Hispanic men aged 25 to 34 years (AAPC, 0.7%; 95% CI, 0.3%-1.1%). By US region, liver cancer mortality rates increased significantly in the West for both Hispanic men (AAPC, 1.6%; 95% CI, 0.9%-2.2%) and Hispanic women (AAPC, 1.5%; 95% CI, 1.1%-1.9%). There were differential findings in mortality rates when comparing Hispanic individuals with individuals belonging to other racial and ethnic populations. Conclusions and Relevance: In this cross-sectional study, despite overall CSM decreasing over 2 decades among Hispanic individuals, disaggregation of data demonstrated that rates of liver cancer deaths among Hispanic men and women and pancreas and uterine cancer deaths among Hispanic women increased from 1999 to 2020. There were also disparities in CSM rates among age groups and US regions. The findings suggest that sustainable solutions need to be implemented to reverse these trends among Hispanic populations.

Longitudinal Analysis of Bladder Cancer-Specific Mortality Trends in the United States.

BACKGROUND: Bladder cancer is the tenth leading cause of cancer death in the United States (US). Advances in diagnosis, imaging, and treatments have led to improvements in bladder cancer management. OBJECTIVE: To evaluate longitudinal bladder cancer mortality trends from 1999-2020 in the US by gender, race, ethnicity, age, geographic region, and urbanization category. METHODS: Age-adjusted bladder cancer death and incidence rates of individuals in the US of all ages between 1999-2020 were obtained using the CDC WONDER and NAACCR databases. Trends and average annual percent changes (AAPC) in age-adjusted Bladder Cancer-Specific Mortality (BCSM) and incidence rates were estimated. Data were analyzed from May 2023 to October 2023. RESULTS: From 1999-2020, overall BCSM decreased by 0.4% annually, with a dramatic decrease in deaths between 2015-2020 (AAPC: -2.0% [95% CI: -2.6,-1.3]). However, BCSM rates and metastatic malignant bladder cancer incidence rates from 1999-2020 increased for individuals≥85 years old (AAPC for BCSM: 0.8% [95% CI:0.5,1.1]; AAPC for metastatic malignant incidence: 2.5% [95% CI: 2.0,2.9]). Increases in BCSM were found for certain years in the South, in rural areas, and for Non-Hispanic White and Asian or Pacific Islander individuals. CONCLUSIONS: Overall mortality from bladder cancer has been decreasing in the US over two decades. Upon disaggregation, increasing trends were found for BCSM and for metastatic malignant bladder cancer incidence for individuals≥85 years old from 1999-2020. Further evaluation of these trends is essential to understand how to target specific populations to improve patient outcomes.

Definitive high-dose, proton-based radiation for unresected mobile spine and sacral chordomas.

BACKGROUND/PURPOSE: Treatment of spine and sacral chordoma generally involves surgical resection, usually in conjunction with radiation therapy.In certain locations, resection may result in significant neurological dysfunction, so definitive radiation has been used as an alternative to surgery. The purpose of this study is to report the results of high-dose, proton-based definitive radiotherapy for unresected spinal and sacral chordomas. MATERIALS/METHODS: Retrospective review of 67 patients with newly diagnosed, unresected spinal chordomas treated with high-dose definitive, proton-based radiotherapy at our center from 1975 to 2019. RESULTS: Reasons for radiotherapy alone included medical inoperability and/or concern for neurological dysfunction based on spine level or patient choice. Tumor locations included cervical (n = 10), thoracic (n = 1), lumbar (n = 4) spine, and sacrum (n = 52). Median maximal tumor diameter was 7.4 cm (range 1.8-25 cm). Median total dose was 77.4 Gy (RBE) (range 73.8-85.9 Gy RBE). Analysis with median follow-up of 56.2 months (range, 4-171.7 months) showed overall survival of 83.5 % (95%CI: 69.4-91.5%) and 65.9% (95%CI: 47.3-79.3%), disease-free survival of 64% (95%CI: 49.3-75.4) and 44.1% (95%CI: 27.8-59.2%), local control of 81.8% (95%CI: 67.6-90.2%) and 63.6% (95%CI: 44.7-77.5%), and distant control of 77.4% (95%CI: 63.6-86.5%) and 72.5% (95%CI: 55.7-83.8%) at 5 and 8 years respectively. The most common late side effect was insufficiency fracture. CONCLUSION: These results continue to support the use of high-dose definitive radiotherapy for patients with medically inoperable or otherwise unresected mobile spine or sacrococcygeal chordomas. There is a trend towards better disease-free survival with doses > 78 Gy (RBE).

Risk of Developing Breast Reconstruction Complications: A Machine-Learning Nomogram for Individualized Risk Estimation with and without Postmastectomy Radiation Therapy.

BACKGROUND: The purpose of this study was to create a nomogram using machine learning models predicting risk of breast reconstruction complications with or without postmastectomy radiation therapy. METHODS: Between 1997 and 2017, 1617 breast cancer patients undergoing mastectomy and breast reconstruction were analyzed. Those with autologous, tissue expander/implant, and single-stage direct-to-implant reconstruction were included. Postmastectomy radiation therapy was delivered either with three-dimensional conformal photon or proton therapy. Complication endpoints were defined based on surgical reintervention operative notes as infection/necrosis requiring débridement. For implant-based patients, complications were defined as capsular contracture requiring capsulotomy and implant failure. For each complication endpoint, least absolute shrinkage and selection operator-penalized regression was used to select the subset of predictors associated with the smallest prediction error from 10-fold cross-validation. Nomograms were built using the least absolute shrinkage and selection operator-selected predictors, and internal validation using cross-validation was performed. RESULTS: Median follow-up was 6.6 years. Among 1617 patients, 23 percent underwent autologous reconstruction, 39 percent underwent direct-to-implant reconstruction, and 37 percent underwent tissue expander/implant reconstruction. Among 759 patients who received postmastectomy radiation therapy, 8.3 percent received proton-therapy to the chest wall and nodes and 43 percent received chest wall boost. Internal validation for each model showed an area under the receiver operating characteristic curve of 73 percent for infection, 75 percent for capsular contracture, 76 percent for absolute implant failure, and 68 percent for overall implant failure. Periareolar incisions and complete implant muscle coverage were found to be important predictors for infection and capsular contracture, respectively. In a multivariable analysis, we found that protons compared to no postmastectomy radiation therapy significantly increased capsular contracture risk (OR, 15.3; p < 0.001). This was higher than the effect of photons with electron boost versus no postmastectomy radiation therapy (OR, 2.5; p = 0.01). CONCLUSION: Using machine learning, these nomograms provided prediction of postmastectomy breast reconstruction complications with and without radiation therapy. CLINICAL QUESTION/LEVEL OF EVIDENCE: Risk, III.

Brain-Specific Relative Biological Effectiveness of Protons Based on Long-term Outcome of Patients With Nasopharyngeal Carcinoma.

PURPOSE: Uncertainties in relative biological effectiveness (RBE) constitute a major pitfall of the use of protons in clinics. An RBE value of 1.1, which is based on cell culture and animal models, is currently used in clinical proton planning. The purpose of this study was to determine RBE for temporal lobe radiographic changes using long-term follow-up data from patients with nasopharyngeal carcinoma. METHODS AND MATERIALS: Five hundred sixty-six patients with newly diagnosed nasopharyngeal carcinoma received double-scattering proton therapy or intensity modulated radiation therapy at our institutions. The 2 treatment cohorts were well matched. Proton dose distributions were simulated using Monte Carlo and compared with those obtained from the proton clinical treatment planning system. Late treatment effect was defined as development of enhancement of temporal lobe on T1-weighted magnetic resonance imaging, with or without accompanying clinical symptoms. The tolerance dose was calculated with receiving operator characteristic analysis and the Youden index. Tolerance curves, expressed as a cumulative dose-volume histogram, were generated using the cutoff points. RESULTS: With a median follow-up period >5 years for both cohorts, 10% of proton patients and 4% of patients undergoing intensity modulated radiation therapy developed temporal lobe enhancement in unilateral temporal lobe. There was no significant difference in dose distributions between the Monte Carlo method and treatment planning system. The tolerance dose-volume levels were V10 (26.1%), V20 (21.9%), V30 (14.0%), V40 (7.7%), V50 (4.8%), and V60 (3.3%) for proton therapy (P < .03). Comparison of the two tolerance curves revealed that tolerance doses of proton treatments were lower than that of photon treatments at all dose levels. The dose tolerance at D1% was 58.56 Gy for protons and 69.07 Gy for photons. The RBE for temporal lobe enhancement from proton treatments were calculated to be 1.18. CONCLUSIONS: Using long-term clinical outcome of patients with nasopharyngeal carcinoma, our data suggest that the RBE for temporal lobe enhancement is 1.18 at D1%. A prospective study in a large cohort would be necessary to confirm these findings.

"Summary Page": a novel tool that reduces omitted data in research databases.

BACKGROUND: Data entry errors are common in clinical research databases. Omitted data are of particular concern because they are more common than erroneously inserted data and therefore could potentially affect research findings. However, few affordable strategies for their prevention are available. METHODS: We have conducted a prospective observational study of the effect of a novel tool called "Summary Page" on the frequency of correction of omitted data errors in a radiation oncology research database between July 2008 and March 2009. "Summary Page" was implemented as an optionally accessed screen in the database that visually integrates key fields in the record. We assessed the frequency of omitted data on the example of the Date of Relapse field. We considered the data in this field to be omitted for all records that had empty Date of Relapse field and evidence of relapse elsewhere in the record. RESULTS: A total of 1,156 records were updated and 200 new records were entered in the database over the study period. "Summary Page" was accessed for 44% of all updated records and for 69% of newly entered records. Frequency of correction of the omitted date of cancer relapse was six-fold higher in records for which "Summary Page" was accessed (p = 0.0003). CONCLUSIONS: "Summary Page" was strongly associated with an increased frequency of correction of omitted data errors. Further, controlled, studies are needed to confirm this finding and elucidate its mechanism of action.