Predicting leptomeningeal disease spread after resection of brain metastases using machine learning.

OBJECTIVE: The incidence of leptomeningeal disease (LMD) has increased as treatments for brain metastases (BMs) have improved and patients with metastatic disease are living longer. Sample sizes of individual studies investigating LMD after surgery for BMs and its risk factors have been limited, ranging from 200 to 400 patients at risk for LMD, which only allows the use of conventional biostatistics. Here, the authors used machine learning techniques to enhance LMD prediction in a cohort of surgically treated BMs. METHODS: A conditional survival forest, a Cox proportional hazards model, an extreme gradient boosting (XGBoost) classifier, an extra trees classifier, and logistic regression were trained. A synthetic minority oversampling technique (SMOTE) was used to train the models and handle the inherent class imbalance. Patients were divided into an 80:20 training and test set. Fivefold cross-validation was used on the training set for hyperparameter optimization. Patients eligible for study inclusion were adults who had consecutively undergone neurosurgical BM treatment, had been admitted to Brigham and Women's Hospital from January 2007 through December 2019, and had a minimum of 1 month of follow-up after neurosurgical treatment. RESULTS: A total of 1054 surgically treated BM patients were included in this analysis. LMD occurred in 168 patients (15.9%) at a median of 7.05 months after BM diagnosis. The discrimination of LMD occurrence was optimal using an XGboost algorithm (area under the curve = 0.83), and the time to LMD was prognosticated evenly by the random forest algorithm and the Cox proportional hazards model (C-index = 0.76). The most important feature for both LMD classification and regression was the BM proximity to the CSF space, followed by a cerebellar BM location. Lymph node metastasis of the primary tumor at BM diagnosis and a cerebellar BM location were the strongest risk factors for both LMD occurrence and time to LMD. CONCLUSIONS: The outcomes of LMD patients in the BM population are predictable using SMOTE and machine learning. Lymph node metastasis of the primary tumor at BM diagnosis and a cerebellar BM location were the strongest LMD risk factors.

Clinical validation of deep learning algorithms for radiotherapy targeting of non-small-cell lung cancer: an observational study.

BACKGROUND: Artificial intelligence (AI) and deep learning have shown great potential in streamlining clinical tasks. However, most studies remain confined to in silico validation in small internal cohorts, without external validation or data on real-world clinical utility. We developed a strategy for the clinical validation of deep learning models for segmenting primary non-small-cell lung cancer (NSCLC) tumours and involved lymph nodes in CT images, which is a time-intensive step in radiation treatment planning, with large variability among experts. METHODS: In this observational study, CT images and segmentations were collected from eight internal and external sources from the USA, the Netherlands, Canada, and China, with patients from the Maastro and Harvard-RT1 datasets used for model discovery (segmented by a single expert). Validation consisted of interobserver and intraobserver benchmarking, primary validation, functional validation, and end-user testing on the following datasets: multi-delineation, Harvard-RT1, Harvard-RT2, RTOG-0617, NSCLC-radiogenomics, Lung-PET-CT-Dx, RIDER, and thorax phantom. Primary validation consisted of stepwise testing on increasingly external datasets using measures of overlap including volumetric dice (VD) and surface dice (SD). Functional validation explored dosimetric effect, model failure modes, test-retest stability, and accuracy. End-user testing with eight experts assessed automated segmentations in a simulated clinical setting. FINDINGS: We included 2208 patients imaged between 2001 and 2015, with 787 patients used for model discovery and 1421 for model validation, including 28 patients for end-user testing. Models showed an improvement over the interobserver benchmark (multi-delineation dataset; VD 0·91 [IQR 0·83-0·92], p=0·0062; SD 0·86 [0·71-0·91], p=0·0005), and were within the intraobserver benchmark. For primary validation, AI performance on internal Harvard-RT1 data (segmented by the same expert who segmented the discovery data) was VD 0·83 (IQR 0·76-0·88) and SD 0·79 (0·68-0·88), within the interobserver benchmark. Performance on internal Harvard-RT2 data segmented by other experts was VD 0·70 (0·56-0·80) and SD 0·50 (0·34-0·71). Performance on RTOG-0617 clinical trial data was VD 0·71 (0·60-0·81) and SD 0·47 (0·35-0·59), with similar results on diagnostic radiology datasets NSCLC-radiogenomics and Lung-PET-CT-Dx. Despite these geometric overlap results, models yielded target volumes with equivalent radiation dose coverage to those of experts. We also found non-significant differences between de novo expert and AI-assisted segmentations. AI assistance led to a 65% reduction in segmentation time (5·4 min; p<0·0001) and a 32% reduction in interobserver variability (SD; p=0·013). INTERPRETATION: We present a clinical validation strategy for AI models. We found that in silico geometric segmentation metrics might not correlate with clinical utility of the models. Experts' segmentation style and preference might affect model performance. FUNDING: US National Institutes of Health and EU European Research Council.

Analysis of Spatial Dose-Volume Relationships and Decline in Sexual Function Following Permanent Brachytherapy for Prostate Cancer.

OBJECTIVE: To explore relationships between dose to periprostatic anatomic structures and erectile dysfunction (ED) outcomes in an institutional cohort treated with prostate brachytherapy. METHODS: The Sexual Health Inventory for Men (SHIM) instrument was administered for stage cT1-T2 prostate cancer patients treated with Pd-103 brachytherapy over a 10-year interval. Dose volume histograms for regional organs at risk and periprostatic regions were calculated with and without expansions to account for contouring uncertainty. Regression tree analysis clustered patients into ED risk groups. RESULTS: We identified 115 men treated with definitive prostate brachytherapy who had 2 years of complete follow-up. On univariate analysis, the subapical region (SAR) caudal to prostate was the only defined region with dose volume histograms parameters significant for potency outcomes. Regression tree analysis separated patients into low ED risk (mean 2-year SHIM 20.03), medium ED risk (15.02), and high ED risk (5.54) groups. Among patients with good baseline function (SHIM ≥ 17), a dose ≥72.75 Gy to 20% of the SAR with 1 cm expansion was most predictive for 2-year potency outcome. On multivariate analysis, regression tree risk group remained significant for predicting potency outcomes even after adjustment for baseline SHIM and age. CONCLUSION: Dose to the SAR immediately caudal to prostate was predictive for potency outcomes in patients with good baseline function. Minimization of dose to this region may improve potency outcomes following prostate brachytherapy.

Adjuvant radiotherapy and outcomes of presumed hemorrhagic melanoma brain metastases without malignant cells.

BACKGROUND: Patients with melanoma can present with a hemorrhagic intracranial lesion. Upon resection, pathology reports may not detect any malignant cells. However, the hemorrhage may obscure their presence and so physicians may still decide whether adjuvant radiotherapy should be applied. Here, we report on the outcomes of a series of patients with melanoma with hemorrhagic brain lesions that returned with no tumor cells. METHODS: All melanoma patients who had craniotomies from 2008 to 2017 at a single institution for hemorrhagic brain lesions were identified through retrospective chart review. Those who had pathology reports with no malignant cells were analyzed. Recurrence at the former site of hemorrhage and resection was the primary outcome. RESULTS: Ten patients met inclusion criteria, and the median follow-up time was 8.5 (1.8-27.3) months. At the time of craniotomy, the median number of brain lesions was 3 (1-25). Two patients had prior craniotomies, eight had prior radiation, and six had prior immunotherapy to the lesion of interest. After surgery, one patient received stereotactic radiosurgery (SRS) to the resection bed. Only one patient developed subsequent melanoma at the resection site; this patient developed the lesion recurrence once and had not received postoperative SRS. CONCLUSION: Although small foci of metastatic disease as a source of bleeding for some patients cannot be excluded, melanoma patients with a suspected hemorrhagic brain metastasis that shows no tumor cells on pathology may benefit from close observation. The local recurrence risk in such cases appears to be low, even without adjuvant radiation.

Prospective evaluation of patient reported swallow function with the Functional Assessment of Cancer Therapy (FACT), MD Anderson Dysphagia Inventory (MDADI) and the Sydney Swallow Questionnaire (SSQ) in head and neck cancer patients.

OBJECTIVES: The Functional Assessment of Cancer Therapy (FACT) instrument is comprised of a group of related and overlapping quality of life (QoL) questionnaires including a core general form, head and neck cancer (HNC)-specific items, and an expert-selected index (FACT-HNSI). Understanding how these relate to more HNC-specific instruments such as the MD Anderson Dysphagia Inventory (MDADI) and Sydney Swallow Questionnaire (SSQ) is vital for guiding their use in clinical trials. MATERIALS AND METHODS: HNC patients concurrently completed MDADI, SSQ, and FACT questionnaires at radiation oncology clinic visits (2015-2016). Spearman correlation coefficients were calculated between each FACT instrument and MDADI or SSQ. Unsupervised k-means cluster analyses were performed to identify clusters of similar QoL responses. Principal component analysis (PCA) identified the degree of variability explained by each instrument. RESULTS: We identified 631 instances (363 patients) where the questionnaires were completed concurrently. Correlations between the various FACT measures and SSQ or MDADI were all significant (p < 0.001), but FACT HNC-specific subscale and FACT-HNSI showed the strongest correlation with MDADI and SSQ. Clustering identified 3 distinct groups of responses when combining instruments either pairwise or three-way. PCA revealed that MDADI and FACT HNC-specific subscale provide similar and likely redundant information. CONCLUSION: FACT HNC-subscale and FACT-HNSI may be preferable over other FACT measures for use in clinical trials where patient-reported swallow function is evaluated. MDADI and FACT provide similar insights into HNC patient QoL while SSQ provides additional, complementary information which could serve to better stratify patients into groups with high, medium, and low QoL outcomes.

Effects of perineural invasion on biochemical recurrence and prostate cancer-specific survival in patients treated with definitive external beam radiotherapy.

OBJECTIVES: Perineural invasion (PNI) has not yet gained universal acceptance as an independent predictor of adverse outcomes for prostate cancer treated with external beam radiotherapy (EBRT). We analyzed the prognostic influence of PNI for a large institutional cohort of prostate cancer patients who underwent EBRT with and without androgen deprivation therapy (ADT). MATERIAL AND METHODS: We, retrospectively, reviewed prostate cancer patients treated with EBRT from 1993 to 2007 at our institution. The primary endpoint was biochemical failure-free survival (BFFS), with secondary endpoints of metastasis-free survival (MFS), prostate cancer-specific survival (PCSS), and overall survival (OS). Univariate and multivariable Cox proportional hazards models were constructed for all survival endpoints. Hazard ratios for PNI were analyzed for the entire cohort and for subsets defined by NCCN risk level. Additionally, Kaplan-Meier survival curves were generated for all survival endpoints after stratification by PNI status, with significant differences computed using the log-rank test. RESULTS: Of 888 men included for analysis, PNI was present on biopsy specimens in 187 (21.1%). PNI was associated with clinical stage, pretreatment PSA level, biopsy Gleason score, and use of ADT (all P<0.01). Men with PNI experienced significantly inferior 10-year BFFS (40.0% vs. 57.8%, P = 0.002), 10-year MFS (79.7% vs. 89.0%, P = 0.001), and 10-year PCSS (90.9% vs. 95.9%, P = 0.009), but not 10-year OS (67.5% vs. 77.5%, P = 0.07). On multivariate analysis, PNI was independently associated with inferior BFFS (P<0.001), but not MFS, PCSS, or OS. In subset analysis, PNI was associated with inferior BFFS (P = 0.04) for high-risk patients and with both inferior BFFS (P = 0.01) and PCSS (P = 0.05) for low-risk patients. Biochemical failure occurred in 33% of low-risk men with PNI who did not receive ADT compared to 8% for low-risk men with PNI treated with ADT (P = 0.01). CONCLUSION: PNI was an independently significant predictor of adverse survival outcomes in this large institutional cohort, particularly for patients with NCCN low-risk disease. PNI should be carefully considered along with other standard prognostic factors when treating these patients with EBRT. Supplementing EBRT with ADT may be beneficial for select low-risk patients with PNI though independent validation with prospective studies is recommended.

Surveillance, epidemiology, and end results-based analysis of the impact of preoperative or postoperative radiotherapy on survival outcomes for T3N0 rectal cancer.

PURPOSE: Preoperative chemoradiation has been established as standard of care for T3/T4 node-positive rectal cancer. Recent work, however, has called into question the overall benefit of radiation for tumors with lower risk characteristics, particularly T3N0 rectal cancers. We retrospectively analyzed T3N0 rectal cancer patients and examined how outcomes differed according to the sequence of treatment received. METHODS: The Surveillance, Epidemiology, and End Results (SEER) database was used to analyze T3N0 rectal cancer cases diagnosed between 1998 and 2008. Treatment consisted of surgery alone (No RT), preoperative radiation followed by surgery (Neo-Adjuvant RT), or surgery followed by postoperative radiation (Adjuvant RT). Demographic and tumor characteristics of the three groups were compared using t-tests for the comparison of means. Survival information from the SEER database was utilized to estimate cause-specific survival (CSS) and to generate Kaplan-Meier survival curves. Multivariate analysis (MVA) of features associated with outcomes was conducted using Cox proportional hazards regression models with Adjuvant RT, Neo-Adjuvant RT, No RT, histological grade, tumor size, year of diagnosis, and demographic characteristics as covariates. RESULTS: 10-Year CSS estimates were 66.1% (95% CI 62.3-69.6%; P=0.02), 73.5% (95% CI 68.9-77.5%; P=0.02), and 76.1% (95% CI 72.4-79.4%; P=0.02), for No RT, Neo-Adjuvant RT, and Adjuvant RT, respectively. On MVA, Adjuvant RT (HR=0.688; 95% CI, 0.578-0.819; P<0.001) was associated with significantly decreased risk for cancer death. By contrast, Neo-Adjuvant RT was not significantly associated with improved cancer survival (HR=0.863; 95% CI, 0.715-1.043; P=0.127). CONCLUSION: Adjuvant RT was associated with significantly higher CSS when compared with surgery alone, while the benefit of Neo-Adjuvant RT was not significant. This indicates that surgery followed by Adjuvant RT may still be an important treatment plan for T3N0 rectal cancer with potentially significant survival advantages over other treatment sequences.

An integrative probabilistic model for identification of structural variation in sequencing data.

Paired-end sequencing is a common approach for identifying structural variation (SV) in genomes. Discrepancies between the observed and expected alignments indicate potential SVs. Most SV detection algorithms use only one of the possible signals and ignore reads with multiple alignments. This results in reduced sensitivity to detect SVs, especially in repetitive regions. We introduce GASVPro, an algorithm combining both paired read and read depth signals into a probabilistic model which can analyze multiple alignments of reads. GASVPro outperforms existing methods with a 50-90% improvement in specificity on deletions and a 50% improvement on inversions.