Faculty Bibliography

PURPOSE:Photon involved-field radiotherapy (IFRT) is the standard-of-care radiotherapy for patients with leptomeningeal metastasis (LM) from solid tumors. We tested whether proton craniospinal irradiation (pCSI) encompassing the entire CNS would result in superior CNS progression-free survival (PFS) compared with IFRT. PATIENTS AND METHODS:We conducted a randomized, phase II trial of pCSI versus IFRT in patients with non-small-cell lung cancer and breast cancers with LM. We enrolled patients with other solid tumors to an exploratory pCSI group. For the randomized groups, patients were assigned (2:1), stratified by histology and systemic disease status, to pCSI or IFRT. The primary end point was CNS PFS. Secondary end points included overall survival (OS) and treatment-related adverse events (TAEs). RESULTS:= .19). In the exploratory pCSI group, 35 patients enrolled, the median CNS PFS was 5.8 months (95% CI, 4.4 to 9.1 months) and OS was 6.6 months (95% CI, 5.4 to 11 months). CONCLUSION:Compared with photon IFRT, we found pCSI improved CNS PFS and OS for patients with non-small-cell lung cancer and breast cancer with LM with no increase in serious TAEs.

Background Artificial intelligence (AI) applications for cancer imaging conceptually begin with automated tumor detection, which can provide the foundation for downstream AI tasks. However, supervised training requires many image annotations, and performing dedicated post hoc image labeling is burdensome and costly. Purpose To investigate whether clinically generated image annotations can be data mined from the picture archiving and communication system (PACS), automatically curated, and used for semisupervised training of a brain MRI tumor detection model. Materials and Methods In this retrospective study, the cancer center PACS was mined for brain MRI scans acquired between January 2012 and December 2017 and included all annotated axial T1 postcontrast images. Line annotations were converted to boxes, excluding boxes shorter than 1 cm or longer than 7 cm. The resulting boxes were used for supervised training of object detection models using RetinaNet and Mask region-based convolutional neural network (R-CNN) architectures. The best-performing model trained from the mined data set was used to detect unannotated tumors on training images themselves (self-labeling), automatically correcting many of the missing labels. After self-labeling, new models were trained using this expanded data set. Models were scored for precision, recall, and F₁ using a held-out test data set comprising 754 manually labeled images from 100 patients (403 intra-axial and 56 extra-axial enhancing tumors). Model F₁ scores were compared using bootstrap resampling. Results The PACS query extracted 31 150 line annotations, yielding 11 880 boxes that met inclusion criteria. This mined data set was used to train models, yielding F₁ scores of 0.886 for RetinaNet and 0.908 for Mask R-CNN. Self-labeling added 18 562 training boxes, improving model F₁ scores to 0.935 (P < .001) and 0.954 (P < .001), respectively. Conclusion The application of semisupervised learning to mined image annotations significantly improved tumor detection performance, achieving an excellent F₁ score of 0.954. This development pipeline can be extended for other imaging modalities, repurposing unused data silos to potentially enable automated tumor detection across radiologic modalities. © RSNA, 2022 Online supplemental material is available for this article.

PURPOSE/UNASSIGNED:Community-academic partnerships have the potential to improve access to clinical trials for under-represented minority patients who more often receive cancer treatment in community settings. In 2017, the Memorial Sloan Kettering (MSK) Cancer Center began opening investigator-initiated clinical trials in radiation oncology in targeted community-based partner sites with a high potential to improve diverse population accrual. This study evaluates the effectiveness of a set of implementation strategies for increasing overall community-based enrollment and the resulting proportional enrollment of Hispanic patients on trials on the basis of availability in community-based partner sites. METHODS/UNASSIGNED:An interrupted time series analysis evaluating implementation strategies was conducted from April 2018 to September 2021. Descriptive analysis ofHispanic enrollment on investigator-initiated randomized therapeutic radiation trials open at community-based sites was compared with those open only at themain academic center. RESULTS/UNASSIGNED:Overall, 84 patients were enrolled in clinical trials in the MSK Alliance, of which 48 (56%) identified as Hispanic. The quarterly patient enrollment pre- vs postimplementation increased from 1.39 (95% CI, -3.67 to 6.46) to 9.42 (95% CI, 2.05 to 16.78; P5 .017). In the investigator-initiated randomized therapeutic radiation trials open in the MSK Alliance, Hispanic representation was 11.5% and 35.9% in twometastatic trials and 14.2% in a proton versus photon trial. Inmatched trials open only at the main academic center, Hispanic representation was 5.6%, 6.0%, and 4.0%, respectively. CONCLUSION/UNASSIGNED:A combination of practice-level and physician-level strategies implemented at community-based partner sites was associated with increased clinical trial enrollment, which translated to improved Hispanic representation. This supports the role Q:2 of strategic community-academic partnerships in addressing disparities in clinical trial enrollment.

Although cancer is highly heterogeneous, all metastatic cancer is considered American Joint Committee on Cancer (AJCC) Stage IV disease. The purpose of this project was to redefine staging of metastatic cancer. Internal validation of nationally representative patient data from the National Cancer Database (n = 461 357; 2010-2013), and external validation using the Surveillance, Epidemiology and End Results database (n = 106 595; 2014-2015) were assessed using the concordance index for evaluation of survival prediction. A Cox proportional hazards model was used for overall survival by considering identified phenotypes (latent classes) and other confounding variables. Latent class analysis was performed for phenotype identification, where Bayesian information criterion (BIC) and sample-size-adjusted BIC were used to select the optimal number of distinct clusters. Kappa coefficients assessed external cluster validation. Latent class analysis identified five metastatic phenotypes with differences in overall survival (P < .0001): (Stage IVA) nearly exclusive bone-only metastases (n = 59 049, 12.8%; median survival 12.7 months; common in lung, breast and prostate cancers); (IVB) predominant lung metastases (n = 62 491, 13.5%; 11.4 months; common in breast, stomach, kidney, ovary, uterus, thyroid, cervix and soft tissue cancers); (IVC) predominant liver/lung metastases (n = 130 014, 28.2%; 7.0 months; common in colorectum, pancreatic, lung, esophagus and stomach cancers); (IVD) bone/liver/lung metastases predominant over brain (n = 61 004, 13.2%; 5.9 months; common in lung and breast cancers); and (IVE) brain/lung metastases predominant over bone/liver (n = 148 799, 32.3%; 5.7 months; lung cancer and melanoma). Long-term survivors were identified, particularly in Stages IVA-B. A pan-cancer nomogram model to predict survival (STARS: site, tumor, age, race, sex) was created, validated and provides 13% better prognostication than AJCC: 1-month concordance index of 0.67 (95% confidence interval [CI]: 0.66-0.67) vs 0.61 (95% CI: 0.60-0.61). STARS is simple, uses easily accessible variables, better prognosticates survival outcomes and provides a platform to develop novel metastasis-directed clinical trials.

BACKGROUND:/Objectives: Pancreatic adenocarcinoma (PDAC) metastatic to the leptomeninges is a rare and lethal event. Leptomeningeal disease (LMD) research is limited in PDAC, and insights into clinical descriptors, possible disease predictors, and treatment strategies is necessitated. METHODS:Memorial Sloan Kettering databases were queried with Institutional Review Board approval to identify patients with LMD and PDAC treated between January 2000 and June 2020. Medical record review was used to abstract clinical, genomic, pathologic, and radiographic data. Overall survival was calculated from date of PDAC diagnosis to date of death. Previously published literature on LMD from PDAC was reviewed. RESULTS:Four patients with LMD from PDAC were identified, two males and two females. Age at diagnosis ranged from 57 to 68 years. All four patients had predominant lung metastasis and a relatively low burden of intra-abdominal disease. Somatic testing indicated alterations typical of PDAC and no PDAC defining pathogenic germline mutations were identified. An extended clinical course prior to LMD diagnosis was observed in all patients, ranging from 16 to 148 months. Upon diagnosis of LMD, three patients elected for supportive care and one patient received a limited course of craniospinal radiation. The median survival following diagnosis of LMD was 1.6 months (range 0.5-2.8 months). CONCLUSIONS:LMD from PDAC is a rare occurrence that may be more frequent in patients with lung metastasis and/or a more indolent clinical course. Following diagnosis of LMD, prognosis is poor, and survival is short. New treatment strategies for this manifestation of PDAC are needed.

CONTEXT/BACKGROUND:Aggressive pituitary tumors that have progressed following temozolomide have limited treatment options. Peptide receptor radionuclide therapy and immunotherapy may have a complementary role in the management of these tumors. METHODS:We provide follow-up data on a previously reported patient with a hypermutated recurrent tumor. The patient in this report provided written informed consent for tumor sequencing and review of medical records on an institutional review board-approved research protocol (NCT01775072). RESULTS:Lu-DOTATATE, the patient had a partial response with a 61% reduction in volume of the target lesion. CONCLUSION/CONCLUSIONS:Ga-DOTATATE positron emission tomography.

PURPOSE:We examined the characteristics of breast cancer patients with oligometastases (OM) treated with stereotactic ablative body radiotherapy (SABR) to identify factors associated with local progression, distant metastasis progression, time to subsequent therapy, progression-free survival (PFS), and overall survival (OS). METHODS:We retrospectively reviewed a single-institution database of patients treated with radiotherapy between 2008 and 2018 and identified 79 patients who received SABR to OM. Twenty-seven patients had genetic testing of metastatic tumors using an institutional targeted sequencing platform. Kaplan-Meier analysis, Cox regression, and competing risk models were used to compare clinical and genetic correlates with outcomes. RESULTS:Median follow-up was 50 months (IQR: 29-66) with 67% of patients alive at the last follow-up. Of the 65% of patients who progressed, 82% progressed outside of the radiation field, 18% experienced local failure, and 80% had oligoprogression. Median OS was 86 months (IQR: 29-66), and PFS was 33 months (IQR: 10-38). Less than 5 years from diagnosis to SABR and triple-negative breast cancer (TNBC) were associated with worse OS. Advanced T stage, any prior chemotherapy, and TNBC were associated with worse PFS. Alterations in CEBPB, RB1, TBX3, PTEN, and CDK4 were associated with worse survival outcomes. CONCLUSION:Long-term systemic disease control and survival can be achieved with SABR for oligometastatic breast cancer. Hormone receptor-positive patients with a long disease interval from initial diagnosis and limited systemic progression history may be ideal for SABR to all sites of disease.

PURPOSE:Guidelines recommend short-course (≤10 fractions) external-beam radiation therapy (EBRT) for bone metastases. Stereotactic body radiation therapy (SBRT) may also improve outcomes; however, routine use is not recommended outside clinical trials. We assessed national radiation therapy trends in complex techniques for bone metastases and associated expenditures. METHODS AND MATERIALS:d or Wilcoxon rank sum tests for categorical and continuous variables, respectively. We identified associations with modality, fractionation, and expenditures using multivariable logistic/linear regression. RESULTS:Among 467,781 radiation episodes for 17 cancer diagnoses, the overall proportion of episodes dedicated to bone metastases (9.4%) was stable from 2015 to 2017, although treatments were increasing in the hospital-affiliated outpatient setting (P < .005). We identified 40,993 episodes for bone metastases, of which 63% were short-course EBRT, 24% were long-course EBRT, 7% were SBRT, and 6% were IMRT. Techniques more common in the hospital-affiliated outpatient setting included short-course EBRT (OPD, 69%, vs FREE, 56%) and SBRT (OPD, 9%, vs FREE, 5%). Techniques more common among free-standing centers included long-course EBRT (OPD, 19%, vs FREE, 31%) and IMRT (OPD, 4%, vs FREE, 9%). From 2015 to 2017, long-course EBRT decreased by an absolute 8%; short-course EBRT, SBRT, and IMRT increased by 4%, 2.5%, and 1%, respectively. The SBRT/IMRT uptake did not differ by setting (P = .4). Differences in expenditures between SBRT and short-course EBRT decreased by a relative 8% in professional and 12% in technical fees. CONCLUSIONS:Approximately 1 in 4 patients received long-course EBRT, with small reductions in use largely replaced by complex treatment modalities. However, expenditures for complex modalities also decreased over time. As alternative payment models take effect, quality metrics are needed to ensure appropriate, effective, and safe delivery of complex technologies.