A probabilistic analysis of completely excised high-grade soft tissue sarcomas of the extremity: an application of a Bayesian belief network

Development and comparison of 1-year survival models in patients with primary bone sarcomas: External validation of a Bayesian belief network model and creation and external validation of a new gradient boosting machine model

Background:

Bone sarcomas often present late with advanced stage at diagnosis and an according, varying short-term survival. In 2016, Nandra et al. generated a Bayesian belief network model for 1-year survival in patients with bone sarcomas. The purpose of this study is: (1) to externally validate the prior 1-year Bayesian belief network prediction model for survival in patients with bone sarcomas and (2) to develop a gradient boosting machine model using Nandra et al.’s cohort and evaluate whether the gradient boosting machine model outperforms the Bayesian belief network model when externally validated in an independent Danish population cohort.

Material and Methods:

The training cohort comprised 3493 patients newly diagnosed with bone sarcoma from the institutional prospectively maintained database at the Royal Orthopaedic Hospital, Birmingham, UK. The validation cohort comprised 771 patients with newly diagnosed bone sarcoma included from the Danish Sarcoma Registry during January 1, 2000–June 22, 2016. We performed area under receiver operator characteristic curve analysis, Brier score and decision curve analysis to evaluate the predictive performance of the models.

Results:

External validation of the Bayesian belief network 1-year prediction model demonstrated an area under receiver operator characteristic curve of 68% (95% confidence interval, 62%-73%). Area under receiver operator characteristic curve of the gradient boosting machine model demonstrated: 75% (95% confidence interval: 70%-80%), overall model performance by the Brier score was 0.09 (95% confidence interval: 0.077–0.11) and decision curve analysis demonstrated a positive net benefit for threshold probabilities above 0.5. External validation of the developed gradient boosting machine model demonstrated an area under receiver operator characteristic curve of 63% (95% confidence interval: 57%-68%), and the Brier score was 0.14 (95% confidence interval: 0.12–0.16).

Conclusion:

External validation of the 1-year Bayesian belief network survival model yielded a poor outcome based on a Danish population cohort validation. We successfully developed a gradient boosting machine 1-year survival model. The gradient boosting machine did not outperform the Bayesian belief network model based on external validation in a Danish population-based cohort.

Functional evaluation following deltoid muscle resection in patients with soft tissue sarcoma

BACKGROUND

The present study aimed to determine functional outcomes in patients undergoing deltoid muscle resection for soft tissue sarcoma.

METHODS

Between 2002 and 2014, 18 patients with soft tissue sarcoma of the shoulder who underwent wide resection including the deltoid muscle, and were followed up for more than 12 months, were retrospectively included in the study. In all, 11 patients were male and 7 were female. The median age was 59 years, median follow-up duration was 37 months. The extent of resection of deltoid muscle, with or without rotator cuff damage, reconstruction methods, adjuvant therapy, oncological outcomes, and the International Society of Limb Salvage (ISOLS) score as functional outcomes were analyzed.

RESULTS

Six patients underwent total resection, and twelve underwent partial resections of deltoid muscle. The rotator cuff was resected in four patients. Soft tissue reconstruction was performed in 17 patients using a pedicled latissimus dorsi muscle flap. Two local recurrences and three distant metastases occurred during follow-up. Median overall survival was 72 months. The mean ISOLS score was 25.0 points (±4.6points). Univariate analysis revealed that there was no significant difference in ISOLS score regarding the extent of deltoid muscle resection. Multivariate analysis identified only combined resection of the rotator cuff as a significant prognostic factor for poor functional outcomes (P < 0.001).

CONCLUSIONS

The extent of resection of the deltoid muscle might not affect the functional outcomes determined by ISOLS score. If the rotator cuff is resected concurrently, satisfactory functional outcomes might not be obtained.

Can a Bayesian Belief Network Be Used to Estimate 1-year Survival in Patients With Bone Sarcomas?

BACKGROUND

Extremity sarcoma has a preponderance to present late with advanced stage at diagnosis. It is important to know why these patients die early from sarcoma and to predict those at high risk. Currently we have mid- to long-term outcome data on which to counsel patients and support treatment decisions, but in contrast to other cancer groups, very little on short-term mortality. Bayesian belief network modeling has been used to develop decision-support tools in various oncologic diagnoses, but to our knowledge, this approach has not been applied to patients with extremity sarcoma.

QUESTIONS/PURPOSES

We sought to (1) determine whether a Bayesian belief network could be used to estimate the likelihood of 1-year mortality using receiver operator characteristic analysis; (2) describe the hierarchal relationships between prognostic and outcome variables; and (3) determine whether the model was suitable for clinical use using decision curve analysis.

METHODS

We considered all patients treated for primary bone sarcoma between 1970 and 2012, and excluded secondary metastasis, presentation with local recurrence, and benign tumors. The institution's database yielded 3499 patients, of which six (0.2%) were excluded. Data extracted for analysis focused on patient demographics (age, sex), tumor characteristics at diagnosis (size, metastasis, pathologic fracture), survival, and cause of death. A Bayesian belief network generated conditional probabilities of variables and survival outcome at 1 year. A lift analysis determined the hierarchal relationship of variables. Internal validation of 699 test patients (20% dataset) determined model accuracy. Decision curve analysis was performed comparing net benefit (capped at 85.5%) for all threshold probabilities (survival output from model).

RESULTS

We successfully generated a Bayesian belief network with five first-degree associates and describe their conditional relationship with survival after the diagnosis of primary bone sarcoma. On internal validation, the resultant model showed good predictive accuracy (area under the curve [AUC] = 0.767; 95% CI, 0.72-0.83). The factors that predict the outcome of interest, 1-year mortality, in order of relative importance are synchronous metastasis (6.4), patient's age (3), tumor size (2.1), histologic grade (1.8), and presentation with a pathologic fracture (1). Patient's sex, tumor location, and inadvertent excision were second-degree associates and not directly related to the outcome of interest. Decision curve analysis shows that clinicians can accurately base treatment decisions on the 1-year model rather than assuming all patients, or no patients, will survive greater than 1 year. For threshold probabilities less than approximately 0.5, the model is no better or no worse than assuming all patients will survive.

CONCLUSIONS

We showed that a Bayesian belief network can be used to predict 1-year mortality in patients presenting with a primary malignancy of bone and quantified the primary factors responsible for an increased risk of death. Synchronous metastasis, patient's age, and the size of the tumor had the largest prognostic effect. We believe models such as these can be useful as clinical decision-support tools and, when properly externally validated, provide clinicians and patients with information germane to the treatment of bone sarcomas.

CLINICAL RELEVANCE

Bone sarcomas are difficult to treat requiring multidisciplinary input to strategize management. An evidence-based survival prediction can be a powerful adjunctive to clinicians in this scenario. We believe the short-term predictions can be used to evaluate services, with 1-year mortality already being a quality indicator. Mortality predictors also can be incorporated in clinical trials, for example, to identify patients who are least likely to experience the side effects of experimental toxic chemotherapeutic agents.

If your lump is bigger than a golf ball and growing, think Sarcoma

AIM

Only 1 in 100 of primary care consultations regarding new soft tissue lumps (STL) are malignant and are susceptible to a delay in diagnosis. We aimed to generate a Bayesian Belief Network to estimate the likelihood of malignancy in patients to facilitate the initial evaluation of a STL and improve timing and quality of referrals to specialist treatment centres.

METHODS

We evaluated all patients referred with a new STL between 1996 and 2007. Variables investigated focused on patient factors, symptoms and STL characteristics. Relevant data was extracted and coded for statistical analysis.

RESULTS

3018 patients with a STL were assessed, of which 1563 (52%) were benign and 1455 (48%) malignant. The features most conditionally associated with the outcome of interest (Benign or Malignant) are referred to as first-degree associates, and are increasing size, age, size of the lump, and duration of symptoms, in that order. On cross validation, this model demonstrated an AUC of 0.77 (95%C.I. 0.75-0.79).

CONCLUSIONS

For the first time, we have described the hierarchal relationship between factors and created an aide memoire, larger than a golf ball and growing, to trigger referral to tertiary tumor units. Importantly, we found pain to be a poor discriminatory factor. We hope our findings will lead to greater awareness and earlier diagnosis of STL.

Lessons learned from the study of 10,000 patients with soft tissue sarcoma

BACKGROUND AND OBJECTIVE

The management of rare tumors is difficult because of limited information on natural history. Our objective was to describe a long-term comprehensive prospective database with the assumption that with careful attention to patient, predisposing tumor and treatment variables, valuable knowledge could be obtained that could guide management.

METHODS

In July of 1982, we began a prospective database of all adult patients admitted to our institution for a surgical procedure for soft tissue sarcoma. Patients were included if they had primary, locally recurrent or metastatic disease undergoing a surgical procedure.

RESULTS

Over 3 decades, we entered 10,000 patients into our prospective soft tissue sarcoma database. Data were entered on a weekly or biweekly schedule with full participation of a multidisciplinary team and a dedicated sarcoma pathologist. Extensive information is available from this database. In this article, we describe distribution by site, histopathology, sex, size, and grade. We utilize this information along with outcome data for local recurrence, distant recurrence, disease specific, and overall survival. The value of molecular diagnosis is illustrated.

CONCLUSIONS

Continuous prospective long-term databases are important to obtain knowledge particularly for rare tumors. Such data can be a rich resource for the development of prognostic indicators including nomograms and can be analyzed by Bayesian Belief Networks. These long-term data linked to collection of tumor and germ-line tissue at the time of an initial procedure will remain a resource for future decades.

Toward better soft tissue sarcoma staging: building on american joint committee on cancer staging systems versions 6 and 7

BACKGROUND

Based on review of patient data in case conferences over time, we hypothesized that clinically relevant data are omitted in routine soft tissue sarcoma staging.

METHODS

We examined subsets of a prospectively collected single institution soft tissue sarcoma database with respect to criteria of the AJCC versions 6 (2002) and 7 (2010) staging systems and examined their clinical outcomes.

RESULTS

Relapse-free survival decreases with increasing primary tumor size in four categories, versus two categories used in AJCC 6 and 7 staging. Disease-specific survival decreases over three categories. Conversely, omission of tumor depth as a prognostic factor in version 7 appears supported, since tumor depth is not an independent risk factor for disease-specific survival by multivariate analysis. Patients with nodal disease and no other metastases fare better than patients with other metastases, but have inferior outcomes compared with patients with large high-grade tumors without nodal metastasis. Multivariate analysis identified size, site, grade, age, nodal metastatic disease, and other metastatic disease as independent risk factors for disease-specific survival. Versions 6 and 7 criteria are tacit regarding anatomic site and histology for tumors with identical FNCLCC grade.

CONCLUSIONS

Improved patient risk assessment may be achieved by staging using a larger number of size categories. Staging system refinements come at the cost of a larger number of staging categories. Histology or site-specific staging systems, nomograms or Bayesian belief networks may provide more accurate means to assess clinical outcomes.