Dataset Information

Developing an Improved Statistical Approach for Survival Estimation in Bone Metastases Management: The Bone Metastases Ensemble Trees for Survival (BMETS) Model.

ABSTRACT:

Purpose

To determine whether a machine learning approach optimizes survival estimation for patients with symptomatic bone metastases (SBM), we developed the Bone Metastases Ensemble Trees for Survival (BMETS) to predict survival using 27 prognostic covariates. To establish its relative clinical utility, we compared BMETS with 2 simpler Cox regression models used in this setting.

Methods and materials

For 492 bone sites in 397 patients evaluated for palliative radiation therapy (RT) for SBM from January 2007 to January 2013, data for 27 clinical variables were collected. These covariates and the primary outcome of time from consultation to death were used to build BMETS using random survival forests. We then performed Cox regressions as per 2 validated models: Chow's 3-item (C-3) and Westhoff's 2-item (W-2) tools. Model performance was assessed using cross-validation procedures and measured by time-dependent area under the curve (tAUC) for all 3 models. For temporal validation, a separate data set comprised of 104 bone sites treated in 85 patients in 2018 was used to estimate tAUC from BMETS.

Results

Median survival was 6.4 months. Variable importance was greatest for performance status, blood cell counts, recent systemic therapy type, and receipt of concurrent nonbone palliative RT. tAUC at 3, 6, and 12 months was 0.83, 0.81, and 0.81, respectively, suggesting excellent discrimination of BMETS across postconsultation time points. BMETS outperformed simpler models at each time, with respective tAUC at each time of 0.78, 0.76, and 0.74 for the C-3 model and 0.80, 0.78, and 0.77 for the W-2 model. For the temporal validation set, respective tAUC was similarly high at 0.86, 0.82, and 0.78.

Conclusions

For patients with SBM, BMETS improved survival predictions versus simpler traditional models. Model performance was maintained when applied to a temporal validation set. To facilitate clinical use, we developed a web platform for data entry and display of BMETS-predicted survival probabilities.

SUBMITTER: Alcorn SR

PROVIDER: S-EPMC7954525 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundBrain metastases (BM) occur frequently in patients with metastatic kidney cancer and are a significant source of morbidity and mortality. Although historically associated with a poor prognosis, survival outcomes for patients in the modern era are incompletely characterized. In particular, outcomes after adjusting for systemic therapy administration and International Metastatic Renal Cell Carcinoma Database Consortium (IMDC) risk factors are not well-known.Patients and methodsA retrospective database of patients with metastatic renal cell carcinoma (RCC) treated at University of Texas Southwestern Medical Center between 2006 and 2015 was created. Data relevant to their diagnosis, treatment course, and outcomes were systematically collected. Survival was analyzed by the Kaplan-Meier method. Patients with BM were compared with patients without BM after adjusting for the timing of BM diagnosis, either prior to or during first-line systemic therapy. The impact of stratification according to IMDC risk group was assessed.ResultsA total of 56 (28.4%) of 268 patients with metastatic RCC were diagnosed with BM prior to or during first-line systemic therapy. Median overall survival (OS) for systemic therapy-naive patients with BM compared with matched patients without BM was 19.5 versus 28.7 months (P = .0117). When analyzed according to IMDC risk group, the median OS for patients with BM was similar for favorable- and intermediate-risk patients (not reached vs. not reached; and 29.0 vs. 36.7 months; P = .5254), and inferior for poor-risk patients (3.5 vs. 9.4 months; P = .0462). For patients developing BM while on first-line systemic therapy, survival from the time of progression did not significantly differ by presence or absence of BM (11.8 vs. 17.8 months; P = .6658).ConclusionsSurvival rates for patients with BM are significantly better than historical reports. After adjusting for systemic therapy, the survival rates of patients with BM in favorable- and intermediate-risk groups were remarkably better than expected and not statistically different from patients without BM, though this represents a single institution experience, and numbers are modest.

Project description:This article reviews the problem of bone disease in prostate cancer and the evolving role of the novel agent denosumab, a fully human monoclonal antibody that inhibits the receptor activator of nuclear factor-?B ligand, in suppressing bone resorption and offering bone protection in this disease. Prostate cancer frequently metastasizes to bone, and additionally its treatment with androgen deprivation leads to accelerated bone loss resulting in clinically relevant skeletal complications associated with disabling symptoms. Among the bone-targeting therapeutic strategies investigated for the prevention of bone complications, the potent bisphosphonate zoledronic acid has been the most widely used agent for bone protection in the past decade. Denosumab is the first among a new class of osteoclast-targeting agents to show superior efficacy in several clinical scenarios in both prostate and breast cancer, as well as in osteoporosis, but the focus of this review will be on its role in prostate cancer. The safety and efficacy of denosumab versus zoledronic acid was established in a randomized trial, demonstrating a delay in skeletal-related events in metastatic castration-resistant prostate cancer patients. This study led to the approval of denosumab in the US. The chief risks of denosumab were hypocalcemia and osteonecrosis of the jaw. Denosumab was also approved for fracture risk reduction in patients on androgen-deprivation therapy for nonmetastatic prostate cancer. Although denosumab extended bone metastasis-free survival in a Phase III trial in men with castration-resistant nonmetastatic prostate cancer to a statistically significant degree, a Food and Drug Administration committee found that the effect was not sufficiently clinically meaningful for regulatory approval, and the Food and Drug Administration issued a letter concurring with the committee's recommendation. The role of denosumab in prostate cancer will continue to evolve either as monotherapy or in combination with other bone-targeting strategies.

Project description:IntroductionThe incidence of prostate cancer in Japan continues to increase, necessitating the continued development of effective therapies and strategies. Recent advances in treatments have improved the prognosis of metastatic disease and highlighted the importance of treating bone metastases to reduce the incidence of skeletal complications and improve patients' quality of life. With the increasing number of treatment options that have become available, including bone-targeted therapy with the alpha emitter radium-223 dichloride (Ra-223), Japanese clinicians are faced with making difficult decisions on the choice of optimal treatment strategy. In such situations, guidance based on expert opinions can be beneficial.MethodsA panel meeting of 27 Japanese experts in the management of prostate cancer was held to share opinions and to establish consensus recommendations on key clinical questions. Panelists were asked to vote on more than 40 questions pertinent to prostate cancer, and the answers helped guide a comprehensive discussion.ResultsThe panel reached a consensus on key topics related to the optimal treatment strategy for Ra-223 therapy, namely, that patients with symptomatic, metastatic castration-resistant prostate cancer (CRPC) would benefit most from the use of this agent and that this treatment therapy should be provided before chemotherapy. Other topics that achieved consensus included: monitoring for osteoporosis and providing treatment if necessary during androgen deprivation therapy; performing magnetic resonance imaging in the presence of discrepancies in bone scintigram and computed tomography scans; monitoring alkaline phosphatase during CRPC treatment; using osteoclast-targeting in patients with CRPC with bone metastases; and using osteoclast-targeted agents combined with Ra-223.ConclusionThese consensus recommendations and the updated information which became available subsequent to the panel meeting included here provide useful information for clinicians to aid in designing optimal treatment strategies for their patients.FundingBayer Yakuhin Ltd.