Improving shared decision making for lung cancer treatment by developing and validating an open-source web based patient decision aid for stage I-II non-small cell lung cancer

The aim of this study was to develop and evaluate a proof-of-concept open-source individualized Patient Decision Aid (iPDA) with a group of patients, physicians, and computer scientists. The iPDA was developed based on the International Patient Decision Aid Standards (IPDAS). A previously published questionnaire was adapted and used to test the user-friendliness and content of the iPDA. The questionnaire contained 40 multiple-choice questions, and answers were given on a 5-point Likert Scale (1-5) ranging from "strongly disagree" to "strongly agree." In addition to the questionnaire, semi-structured interviews were conducted with patients. We performed a descriptive analysis of the responses. The iPDA was evaluated by 28 computer scientists, 21 physicians, and 13 patients. The results demonstrate that the iPDA was found valuable by 92% (patients), 96% (computer scientists), and 86% (physicians), while the treatment information was judged useful by 92%, 96%, and 95%, respectively. Additionally, the tool was thought to be motivating for patients to actively engage in their treatment by 92%, 93%, and 91% of the above respondents groups. More multimedia components and less text were suggested by the respondents as ways to improve the tool and user interface. In conclusion, we successfully developed and tested an iPDA for patients with stage I-II Non-Small Cell Lung Cancer (NSCLC).

Towards texture accurate slice interpolation of medical images using PixelMiner

Quantitative image analysis models are used for medical imaging tasks such as registration, classification, object detection, and segmentation. For these models to be capable of making accurate predictions, they need valid and precise information. We propose PixelMiner, a convolution-based deep-learning model for interpolating computed tomography (CT) imaging slices. PixelMiner was designed to produce texture-accurate slice interpolations by trading off pixel accuracy for texture accuracy. PixelMiner was trained on a dataset of 7829 CT scans and validated using an external dataset. We demonstrated the model's effectiveness by using the structural similarity index (SSIM), peak signal to noise ratio (PSNR), and the root mean squared error (RMSE) of extracted texture features. Additionally, we developed and used a new metric, the mean squared mapped feature error (MSMFE). The performance of PixelMiner was compared to four other interpolation methods: (tri-)linear, (tri-)cubic, windowed sinc (WS), and nearest neighbor (NN). PixelMiner produced texture with a significantly lowest average texture error compared to all other methods with a normalized root mean squared error (NRMSE) of 0.11 (p < .01), and the significantly highest reproducibility with a concordance correlation coefficient (CCC) ≥ 0.85 (p < .01). PixelMiner was not only shown to better preserve features but was also validated using an ablation study by removing auto-regression from the model and was shown to improve segmentations on interpolated slices.

Automated detection and segmentation of non-small cell lung cancer computed tomography images

Detection and segmentation of abnormalities on medical images is highly important for patient management including diagnosis, radiotherapy, response evaluation, as well as for quantitative image research. We present a fully automated pipeline for the detection and volumetric segmentation of non-small cell lung cancer (NSCLC) developed and validated on 1328 thoracic CT scans from 8 institutions. Along with quantitative performance detailed by image slice thickness, tumor size, image interpretation difficulty, and tumor location, we report an in-silico prospective clinical trial, where we show that the proposed method is faster and more reproducible compared to the experts. Moreover, we demonstrate that on average, radiologists & radiation oncologists preferred automatic segmentations in 56% of the cases. Additionally, we evaluate the prognostic power of the automatic contours by applying RECIST criteria and measuring the tumor volumes. Segmentations by our method stratified patients into low and high survival groups with higher significance compared to those methods based on manual contours.

Correct interpretation of computer tomography (CT) scans is important for the correct assessment of a patient’s disease but can be subjective and timely. Here, the authors develop a system that can automatically segment the non-small cell lung cancer on CT images of patients and show in an in silico trial that the method was faster and more reproducible than clinicians.

AB1469 SPANISH TRANSLATION AND CROSS-CULTURAL ADAPTATION OF THE mSQUASH

Background

Regular physical activity is recommended for all patients in the ASAS/EULAR recommendations for the management of axial spondyloarthritis (axSpA). However, there is a lack of outcome measures that assess the amount and type of physical activity in patients with axSpA. For this matter, the modified Short QUestionnaire to Assess Health enhancing physical activity (mSQUASH) was developed and validated, originally in Dutch1.

Objectives

To translate and cross-culturally adapt the mSQUASH into Spanish and to test the equivalence of the translated version in patients with axSpA.

Methods

The mSQUASH was translated into Spanish and then back-translated into Dutch, following forward-backward procedure as described by Beaton2 (Figure 1). Two bi-lingual translators (native speakers for European Spanish) produced independent forward translations of the item content, response options, and instructions of the mSQUASH into Spanish. Both versions were harmonized in a meeting among the Spanish translators, a methodologist and a rheumatologist into a consensual version. Another translator (native speaker for Dutch), blinded for the original version, back translated the synthesized version into Dutch. An expert committee, including all translators, one methodologist and a rheumatologist, reached consensus on discrepancies, ensuring equivalence between the Dutch and Spanish versions, and developed a pre-final version of the Spanish mSQUASH. The field test with cognitive debriefing involved a sample of 10 patients with axSpA covering the full spectrum of the disease -radiographic axSpA (r-axSpA) and non-radiographic axSpA (nr-axSpA)- with different gender, age, disease duration, and educational background. Each patient was interviewed to check understandability, interpretation and cultural relevance of the translation.

Figure 1.

Cross-cultural adaptation of the mSQUASH

Results

The translation process of the mSQUASH was completed without major complications following the forward-backward procedure. The first translation needed several iterations due to small discrepancies in the wording. Back-translation was performed without difficulties, and the expert committee agreed upon a final version of the questionnaire. A total of 10 patients with axSpA participated in the field test (Table 1). Seven were male, mean age (SD) was 38.9 (14.4) years; 6 patients had r-axSpA, 9 were HLA-B27+. Cognitive debriefing showed the Spanish questionnaire to be, relevant, understandable and comprehensive. The preliminary version was accepted with minor modifications. As a result of the interviews, minor spelling errors were corrected, and the wording of the response categories was homogenized (“despacio/ligero”). Besides, the term “colegio”- translated literally from the Dutch “school”- was found not comprehensive enough to reflect possibilities on education (i.e. it does not include university), so it was adapted to “el lugar de estudio”.

Table 1.

Patients’ characteristics

#GenderAgeWorking statusEducationaxSpA subtypeDisease durationHLA-B27DrugBASDAI1Male63WorkingUniversityr-axSpA35 y+NSAIDs2.32Male24StudentSecondaryr-axSpA6 y+NSAIDs03Male37WorkingUniversityr-axSpA5 y+ADA2.54Male66RetiredUniversityr-axSpA23 y+IFN3.15Male29WorkingUniversityr-axSpA11 y+ADA06Female26WorkingUniversitynr-axSpA2 y+NSAIDs-7Male24StudentUniversitynr-axSpA1 y+ETA4.58Male35WorkingUniversityr-axSpA12 y+GOL09Female40WorkingSecondarynr-axSpA4 y+NSAIDs-10Female45UnemployedPrimarynr-axSpA9 y-GOL8.2

Conclusion

The resulting Spanish version of the mSQUASH showed good linguistic and face validity according to the field test, revealing potential for use in both clinical practice and research settings. In order to conclude the cross-cultural adaptation of the mSQUASH into Spanish, the next step is the assessment of psychometric properties of the Spanish version.

References

[1]Beaton DE, et al. Spine. 2000; 25:3186-91

[2]Carbo et al. Semin Arthritis Rheum. 2021; 51:719-27

Disclosure of Interests

Diego Benavent Speakers bureau: Jannsen, Roche, Grant/research support from: Novartis, Andrea Jochems: None declared, DORA PASCUAL-SALCEDO Speakers bureau: Pfizer, Menarini, Takeda, Abvvie., Grant/research support from: Pfizer, Menarini, Takeda, Abvvie., Gijs Jochems: None declared, Chamaida Plasencia Speakers bureau: Pfizer, Abbvie, Lilly, Sandoz, Sanofi, Biogen, Roche and Novartis, Grant/research support from: Pfizer and Abbvie, Sofia Ramiro Speakers bureau: Eli Lilly, MSD, Novartis, UCB, Consultant of: AbbVie, Eli Lilly, MSD, Novartis, Pfizer, UCB, Sanofi, Grant/research support from: AbbVie, Galapagos, Novartis, Pfizer, UCB, Suzanne Arends: None declared, Anneke Spoorenberg Consultant of: AbbVie, Novartis, Pfizer; UCB, Lilly, Grant/research support from: AbbVie, Pfizer, Alejandro Balsa Speakers bureau: Pfizer, Abbvie, Lilly, Galapagos, BMS, Sandoz, Nordic Pharma, Gebro, Roche, Sanofi, UCB, Consultant of: Pfizer, Abbvie, Lilly, Galapagos, BMS, Nordic Pharma, Sanofi, UCB, Grant/research support from: Pfizer, Abbvie, BMS, Nordic Pharma, Gebro, Roche, UCB, Victoria Navarro-Compán Speakers bureau: AbbVie, Eli Lilly, Janssen, MSD, Novartis, Pfizer, UCB Pharma, Consultant of: AbbVie, Eli Lilly, MSD, Novartis, Pfizer, UCB Pharma, Grant/research support from: AbbVie and Novartis

AB1472 TRANSLATION AND CROSS-CULTURAL ADAPTATION OF THE CORS INTO SPANISH

Background

Rheumatic diseases substantially affect the lives of patients, with complex associations between disease severity and self-perceived health status. In this regard, the Coping with Rheumatic Stressors (CORS) questionnaire was developed to measure how patients with rheumatoid arthritis cope with stressors such as pain or dependence. There is no validated instrument to measure coping in axial spondyloarthritis (axSpA) and therefore the adaptation of the CORS would be of great value.

Objectives

To cross-culturally adapt the CORS into Spanish and to test the conceptual equivalence of the translated version in patients with axSpA.

Methods

A translation of the CORS into Spanish was performed, followed by a back-translation into Dutch, following forward-backward procedure as described by Beaton1(Figure 1). Two bi-lingual translators (native speakers for Spanish), one of them informed of the content of the questionnaire and the other not informed, produced independent forward translations of the item content, response options, and instructions of the CORS into Spanish. Both versions were harmonized in a consensual version. Another translator (native speaker for Dutch), not informed of the concepts used in the questionnaire, back translated the synthesized version into Dutch. An expert committee including all translators, one methodologist and a rheumatologist, held a meeting and reached consensus on discrepancies to develop a pre-final version of the Spanish CORS. The field test with cognitive debriefing involved a sample of 10 patients with axSpA covering the full spectrum of the disease and with different sociodemographic backgrounds.

Figure 1.

Cross-cultural adaptation of the CORS

Results

The translation process of the CORS was completed following the forward-backward procedure, after discussion of the discrepancies throughout the process. The first translation was done without major complications. However, several discrepancies appeared in the back-translation, in which there were minor modifications in the wording in one response option (“muchas veces” to “muy a menudo”) and 15 questionnaire items. As an example, “Ik ga de deur uit”, literally meaning “I go out by the door”, was initially translated as such (“salgo por la puerta”); however, it conceptually represents “I go away”, and it was adapted like this (“me voy a la calle”). Thus, a pre-final consensus version of the CORS was agreed by the expert committee. This pre-final version was field tested in 10 patients with axSpA: mean age (SD) was 38.9 (14.4) years, 7 patients were male, 6 had radiographic axSpA, and 9 were HLA-B27+. The Spanish questionnaire appeared clear and understandable to all patients. However, some minor modifications were proposed in some items (Table 1). As a result of the cognitive debriefing, two changes were implemented (one instruction and one item), whereas two other suggestions did not lead to any change due to minor wording discrepancies with similar conceptual equivalence. The final version of the Spanish CORS is shown at shorturl.at/cimC6.

Table 1.

Cognitive debriefing queries and decisions from the expert committee

Original Dutch itemSpanish translation pre-final# Patient queriesQueriesFinal version(….) aan te geven hoe vaak u het beschreven gedrag uitvoert.(…) indique cuán a menudo usted ha llevado a cabo dicho comportamiento.1Literal discrepancies(…) indique la frecuencia con que usted ha tenido dicho comportamiento.Ik rust op tijd uitMe voy a tiempo a descansar1Literal discrepanciesNo changesIk probeer er het beste van te makenIntento aprovechar al máximo1Literal discrepanciesNo changesIk houd rekening met anderenTengo en cuenta a los demás2Meaning doubtsTengo en consideración a los que me ayudan/cuidan

Conclusion

The Spanish version of the CORS showed good cross-cultural validity and good face validity in patients with axSpA according to the field test. Before the Spanish CORS is implemented, further validation is in progress to test the psychometric properties of the instrument.

References

[1]Beaton DE, et al. Spine. 2000; 25:3186-91

Disclosure of Interests

Diego Benavent Speakers bureau: Jannsen, Roche, Grant/research support from: Novartis, Andrea Jochems: None declared, DORA PASCUAL-SALCEDO Speakers bureau: Pfizer, Menarini, Takeda, Abvvie., Grant/research support from: Pfizer, Menarini, Takeda, Abvvie., Gijs Jochems: None declared, Chamaida Plasencia Speakers bureau: Pfizer, Abvvie, Lilly, Sandoz, Sanofi, Biogen, Roche, Novartis., Grant/research support from: Pfizer, Abvvie., Sofia Ramiro Speakers bureau: Eli Lilly, MSD, Novartis, UC, Consultant of: AbbVie, Eli Lilly, MSD, Novartis, Pfizer, UCB, Sanofi, Grant/research support from: AbbVie, Galapagos, Novartis, Pfizer, UCB, Wim van Lankveld: None declared, Alejandro Balsa Speakers bureau: Pfizer, Abbvie, Lilly, Galapagos, BMS, Sandoz, Nordic Pharma, Gebro, Roche, Sanofi, UCB, Consultant of: Pfizer, Abbvie, Lilly, Galapagos, BMS, Nordic Pharma, Sanofi, UCB, Grant/research support from: Pfizer, Abbvie, BMS, Nordic Pharma, Gebro, Roche, UCB, Victoria Navarro-Compán Speakers bureau: AbbVie, Eli Lilly, Janssen, MSD, Novartis, Pfizer, UCB Pharma, Consultant of: AbbVie, Eli Lilly, MSD, Novartis, Pfizer, UCB Pharma, Grant/research support from: AbbVie and Novartis

Selectively Targeting Tumor Hypoxia With the Hypoxia-Activated Prodrug CP-506

Hypoxia-activated prodrugs (HAP) are a promising class of antineoplastic agents that can selectively eliminate hypoxic tumor cells. This study evaluates the hypoxia-selectivity and antitumor activity of CP-506, a DNA alkylating HAP with favorable pharmacologic properties. Stoichiometry of reduction, one-electron affinity, and back-oxidation rate of CP-506 were characterized by fast-reaction radiolytic methods with observed parameters fulfilling requirements for oxygen-sensitive bioactivation. Net reduction, metabolism, and cytotoxicity of CP-506 were maximally inhibited at oxygen concentrations above 1 μmol/L (0.1% O₂). CP-506 demonstrated cytotoxicity selectively in hypoxic 2D and 3D cell cultures with normoxic/anoxic IC₅₀ ratios up to 203. Complete resistance to aerobic (two-electron) metabolism by aldo-keto reductase 1C3 was confirmed through gain-of-function studies while retention of hypoxic (one-electron) bioactivation by various diflavin oxidoreductases was also demonstrated. In vivo, the antitumor effects of CP-506 were selective for hypoxic tumor cells and causally related to tumor oxygenation. CP-506 effectively decreased the hypoxic fraction and inhibited growth of a wide range of hypoxic xenografts. A multivariate regression analysis revealed baseline tumor hypoxia and in vitro sensitivity to CP-506 were significantly correlated with treatment response. Our results demonstrate that CP-506 selectively targets hypoxic tumor cells and has broad antitumor activity. Our data indicate that tumor hypoxia and cellular sensitivity to CP-506 are strong determinants of the antitumor effects of CP-506.

Modeling-Based Decision Support System for Radical Prostatectomy Versus External Beam Radiotherapy for Prostate Cancer Incorporating an In Silico Clinical Trial and a Cost-Utility Study

Simple Summary

Low–intermediate prostate cancer has a number of viable treatment options, such as radical prostatectomy and radiotherapy, with similar survival outcomes but different treatment-related side effects. The aim of this study is to facilitate patient-specific treatment selection by developing a decision support system (DSS) that incorporates predictive models for cancer-free survival and treatment-related side effects. We challenged this DSS by validating it against randomized clinical trials and assessing the benefit through a cost–utility analysis. We aim to expand upon the applications of this DSS by using it as the basis for an in silico clinical trial for an underrepresented patient group. This modeling study shows that DSS-based treatment decisions will result in a clinically relevant increase in the patients’ quality of life and can be used for in silico trials.

Abstract

The aim of this study is to build a decision support system (DSS) to select radical prostatectomy (RP) or external beam radiotherapy (EBRT) for low- to intermediate-risk prostate cancer patients. We used an individual state-transition model based on predictive models for estimating tumor control and toxicity probabilities. We performed analyses on a synthetically generated dataset of 1000 patients with realistic clinical parameters, externally validated by comparison to randomized clinical trials, and set up an in silico clinical trial for elderly patients. We assessed the cost-effectiveness (CE) of the DSS for treatment selection by comparing it to randomized treatment allotment. Using the DSS, 47.8% of synthetic patients were selected for RP and 52.2% for EBRT. During validation, differences with the simulations of late toxicity and biochemical failure never exceeded 2%. The in silico trial showed that for elderly patients, toxicity has more influence on the decision than TCP, and the predicted QoL depends on the initial erectile function. The DSS is estimated to result in cost savings (EUR 323 (95% CI: EUR 213–433)) and more quality-adjusted life years (QALYs; 0.11 years, 95% CI: 0.00–0.22) than randomized treatment selection.

Structural and functional radiomics for lung cancer

INTRODUCTION

Lung cancer ranks second in new cancer cases and first in cancer-related deaths worldwide. Precision medicine is working on altering treatment approaches and improving outcomes in this patient population. Radiological images are a powerful non-invasive tool in the screening and diagnosis of early-stage lung cancer, treatment strategy support, prognosis assessment, and follow-up for advanced-stage lung cancer. Recently, radiological features have evolved from solely semantic to include (handcrafted and deep) radiomic features. Radiomics entails the extraction and analysis of quantitative features from medical images using mathematical and machine learning methods to explore possible ties with biology and clinical outcomes.

METHODS

Here, we outline the latest applications of both structural and functional radiomics in detection, diagnosis, and prediction of pathology, gene mutation, treatment strategy, follow-up, treatment response evaluation, and prognosis in the field of lung cancer.

CONCLUSION

The major drawbacks of radiomics are the lack of large datasets with high-quality data, standardization of methodology, the black-box nature of deep learning, and reproducibility. The prerequisite for the clinical implementation of radiomics is that these limitations are addressed. Future directions include a safer and more efficient model-training mode, merge multi-modality images, and combined multi-discipline or multi-omics to form "Medomics."

Survival outcomes of patients with advanced melanoma from 2013 to 2017: Results of a nationwide population-based registry

BACKGROUND

The treatment landscape has completely changed for advanced melanoma. We report survival outcomes and the differential impact of prognostic factors over time in daily clinical practice.

METHODS

From a Dutch nationwide population-based registry, patients with advanced melanoma diagnosed from 2013 to 2017 were analysed (n = 3616). Because the proportional hazards assumption was violated, a multivariable Cox model restricted to the first 6 months and a multivariable landmark Cox model from 6 to 48 months were used to assess overall survival (OS) of cases without missing values. The 2017 cohort was excluded from this analysis because of the short follow-up time.

RESULTS

Median OS of the 2013 and 2016 cohort was 11.7 months (95% confidence interval [CI]: 10.4-13.5) and 17.7 months (95% CI: 14.9-19.8), respectively. Compared with the 2013 cohort, the 2016 cohort had superior survival in the Cox model from 0 to 6 months (hazard ratio [HR] = 0.55 [95% CI: 0.43-0.72]) and in the Cox model from 6 to 48 months (HR = 0.68 [95% CI: 0.57-0.83]). Elevated lactate dehydrogenase levels, distant metastases in ≥3 organ sites, brain and liver metastasis and Eastern Cooperative Oncology Group performance score of ≥1 had stronger association with inferior survival from 0 to 6 months than from 6 to 48 months. BRAF-mutated melanoma had superior survival in the first 6 months (HR = 0.50 [95% CI: 0.42-0.59]).

CONCLUSION(S)

Prognosis for advanced melanoma in the Netherlands has improved from 2013 to 2016. Prognostic importance of most evaluated factors was higher in the first 6 months after diagnosis. BRAF-mutated melanoma was only associated with superior survival in the first 6 months.

Non-invasive imaging prediction of tumor hypoxia: A novel developed and externally validated CT and FDG-PET-based radiomic signatures

BACKGROUND

Tumor hypoxia increases resistance to radiotherapy and systemic therapy. Our aim was to develop and validate a disease-agnostic and disease-specific CT (+FDG-PET) based radiomics hypoxia classification signature.

MATERIAL AND METHODS

A total of 808 patients with imaging data were included: N = 100 training/N = 183 external validation cases for a disease-agnostic CT hypoxia classification signature, N = 76 training/N = 39 validation cases for the H&N CT signature and N = 62 training/N = 36 validation cases for the Lung CT signature. The primary gross tumor volumes (GTV) were manually defined by experts on CT. In order to dichotomize between hypoxic/well-oxygenated tumors a threshold of 20% was used for the [¹⁸F]-HX4-derived hypoxic fractions (HF). A random forest (RF)-based machine-learning classifier/regressor was trained to classify patients as hypoxia-positive/ negative based on radiomic features.

RESULTS

A 11 feature "disease-agnostic CT model" reached AUC's of respectively 0.78 (95% confidence interval [CI], 0.62-0.94), 0.82 (95% CI, 0.67-0.96) and 0.78 (95% CI, 0.67-0.89) in three external validation datasets. A "disease-agnostic FDG-PET model" reached an AUC of 0.73 (0.95% CI, 0.49-0.97) in validation by combining 5 features. The highest "lung-specific CT model" reached an AUC of 0.80 (0.95% CI, 0.65-0.95) in validation with 4 CT features, while the "H&N-specific CT model" reached an AUC of 0.84 (0.95% CI, 0.64-1.00) in validation with 15 CT features. A tumor volume-alone model was unable to significantly classify patients as hypoxia-positive/ negative. A significant survival split (P = 0.037) was found between CT-classified hypoxia strata in an external H&N cohort (n = 517), while 117 significant hypoxia gene-CT signature feature associations were found in an external lung cohort (n = 80).

CONCLUSION

The disease-specific radiomics signatures perform better than the disease agnostic ones. By identifying hypoxic patients our signatures have the potential to enrich interventional hypoxia-targeting trials.

Decision Support Systems in Oncology

Precision medicine is the future of health care: please watch the animation at https://vimeo.com/241154708. As a technology-intensive and -dependent medical discipline, oncology will be at the vanguard of this impending change. However, to bring about precision medicine, a fundamental conundrum must be solved: Human cognitive capacity, typically constrained to five variables for decision making in the context of the increasing number of available biomarkers and therapeutic options, is a limiting factor to the realization of precision medicine. Given this level of complexity and the restriction of human decision making, current methods are untenable. A solution to this challenge is multifactorial decision support systems (DSSs), continuously learning artificial intelligence platforms that integrate all available data—clinical, imaging, biologic, genetic, cost—to produce validated predictive models. DSSs compare the personalized probable outcomes—toxicity, tumor control, quality of life, cost effectiveness—of various care pathway decisions to ensure optimal efficacy and economy. DSSs can be integrated into the workflows both strategically (at the multidisciplinary tumor board level to support treatment choice, eg, surgery or radiotherapy) and tactically (at the specialist level to support treatment technique, eg, prostate spacer or not). In some countries, the reimbursement of certain treatments, such as proton therapy, is already conditional on the basis that a DSS is used. DSSs have many stakeholders—clinicians, medical directors, medical insurers, patient advocacy groups—and are a natural consequence of big data in health care. Here, we provide an overview of DSSs, their challenges, opportunities, and capacity to improve clinical decision making, with an emphasis on the utility in oncology.

Deep learning in fracture detection: a narrative review

Artificial intelligence (AI) is a general term that implies the use of a computer to model intelligent behavior with minimal human intervention. AI, particularly deep learning, has recently made substantial strides in perception tasks allowing machines to better represent and interpret complex data. Deep learning is a subset of AI represented by the combination of artificial neuron layers. In the last years, deep learning has gained great momentum. In the field of orthopaedics and traumatology, some studies have been done using deep learning to detect fractures in radiographs. Deep learning studies to detect and classify fractures on computed tomography (CT) scans are even more limited. In this narrative review, we provide a brief overview of deep learning technology: we (1) describe the ways in which deep learning until now has been applied to fracture detection on radiographs and CT examinations; (2) discuss what value deep learning offers to this field; and finally (3) comment on future directions of this technology.

Privacy-preserving distributed learning of radiomics to predict overall survival and HPV status in head and neck cancer

A major challenge in radiomics is assembling data from multiple centers. Sharing data between hospitals is restricted by legal and ethical regulations. Distributed learning is a technique, enabling training models on multicenter data without data leaving the hospitals ("privacy-preserving" distributed learning). This study tested feasibility of distributed learning of radiomics data for prediction of two year overall survival and HPV status in head and neck cancer (HNC) patients. Pretreatment CT images were collected from 1174 HNC patients in 6 different cohorts. 981 radiomic features were extracted using Z-Rad software implementation. Hierarchical clustering was performed to preselect features. Classification was done using logistic regression. In the validation dataset, the receiver operating characteristics (ROC) were compared between the models trained in the centralized and distributed manner. No difference in ROC was observed with respect to feature selection. The logistic regression coefficients were identical between the methods (absolute difference <10-7). In comparison of the full workflow (feature selection and classification), no significant difference in ROC was found between centralized and distributed models for both studied endpoints (DeLong p > 0.05). In conclusion, both feature selection and classification are feasible in a distributed manner using radiomics data, which opens new possibility for training more reliable radiomics models.

Systematic Review of Privacy-Preserving Distributed Machine Learning From Federated Databases in Health Care

Big data for health care is one of the potential solutions to deal with the numerous challenges of health care, such as rising cost, aging population, precision medicine, universal health coverage, and the increase of noncommunicable diseases. However, data centralization for big data raises privacy and regulatory concerns.Covered topics include (1) an introduction to privacy of patient data and distributed learning as a potential solution to preserving these data, a description of the legal context for patient data research, and a definition of machine/deep learning concepts; (2) a presentation of the adopted review protocol; (3) a presentation of the search results; and (4) a discussion of the findings, limitations of the review, and future perspectives.Distributed learning from federated databases makes data centralization unnecessary. Distributed algorithms iteratively analyze separate databases, essentially sharing research questions and answers between databases instead of sharing the data. In other words, one can learn from separate and isolated datasets without patient data ever leaving the individual clinical institutes.Distributed learning promises great potential to facilitate big data for medical application, in particular for international consortiums. Our purpose is to review the major implementations of distributed learning in health care.

Preoperative CT-based radiomics combined with intraoperative frozen section is predictive of invasive adenocarcinoma in pulmonary nodules: a multicenter study

Objectives

Develop a CT-based radiomics model and combine it with frozen section (FS) and clinical data to distinguish invasive adenocarcinomas (IA) from preinvasive lesions/minimally invasive adenocarcinomas (PM).

Methods

This multicenter study cohort of 623 lung adenocarcinomas was split into training (n = 331), testing (n = 143), and external validation dataset (n = 149). Random forest models were built using selected radiomics features, results from FS, lesion volume, clinical and semantic features, and combinations thereof. The area under the receiver operator characteristic curves (AUC) was used to evaluate model performances. The diagnosis accuracy, calibration, and decision curves of models were tested.

Results

The radiomics-based model shows good predictive performance and diagnostic accuracy for distinguishing IA from PM, with AUCs of 0.89, 0.89, and 0.88, in the training, testing, and validation datasets, respectively, and with corresponding accuracies of 0.82, 0.79, and 0.85. Adding lesion volume and FS significantly increases the performance of the model with AUCs of 0.96, 0.97, and 0.96, and with accuracies of 0.91, 0.94, and 0.93 in the three datasets. There is no significant difference in AUC between the FS model enriched with radiomics and volume against an FS model enriched with volume alone, while the former has higher accuracy. The model combining all available information shows minor non-significant improvements in AUC and accuracy compared with an FS model enriched with radiomics and volume.

Conclusions

Radiomics signatures are potential biomarkers for the risk of IA, especially in combination with FS, and could help guide surgical strategy for pulmonary nodules patients.

Key Points

• A CT-based radiomics model may be a valuable tool for preoperative prediction of invasive adenocarcinoma for patients with pulmonary nodules.

• Radiomics combined with frozen sections could help in guiding surgery strategy for patients with pulmonary nodules.

Electronic supplementary material

The online version of this article (10.1007/s00330-019-06597-8) contains supplementary material, which is available to authorized users.

Distributed learning on 20 000+ lung cancer patients - The Personal Health Train

BACKGROUND AND PURPOSE

Access to healthcare data is indispensable for scientific progress and innovation. Sharing healthcare data is time-consuming and notoriously difficult due to privacy and regulatory concerns. The Personal Health Train (PHT) provides a privacy-by-design infrastructure connecting FAIR (Findable, Accessible, Interoperable, Reusable) data sources and allows distributed data analysis and machine learning. Patient data never leaves a healthcare institute.

MATERIALS AND METHODS

Lung cancer patient-specific databases (tumor staging and post-treatment survival information) of oncology departments were translated according to a FAIR data model and stored locally in a graph database. Software was installed locally to enable deployment of distributed machine learning algorithms via a central server. Algorithms (MATLAB, code and documentation publicly available) are patient privacy-preserving as only summary statistics and regression coefficients are exchanged with the central server. A logistic regression model to predict post-treatment two-year survival was trained and evaluated by receiver operating characteristic curves (ROC), root mean square prediction error (RMSE) and calibration plots.

RESULTS

In 4 months, we connected databases with 23 203 patient cases across 8 healthcare institutes in 5 countries (Amsterdam, Cardiff, Maastricht, Manchester, Nijmegen, Rome, Rotterdam, Shanghai) using the PHT. Summary statistics were computed across databases. A distributed logistic regression model predicting post-treatment two-year survival was trained on 14 810 patients treated between 1978 and 2011 and validated on 8 393 patients treated between 2012 and 2015.

CONCLUSION

The PHT infrastructure demonstrably overcomes patient privacy barriers to healthcare data sharing and enables fast data analyses across multiple institutes from different countries with different regulatory regimens. This infrastructure promotes global evidence-based medicine while prioritizing patient privacy.

NIMG-65. PREDICTING PROGNOSIS AND CANCER HOTSPOT MUTATIONS USING QUALITATIVE MR IMAGING ANALYSIS IN GLIOBLASTOMA

INTRODUCTION

Tumor heterogeneity poses one of the major limitations in improving the treatment for glioblastoma (GBM), which calls for new clinically relevant predictive models. This study aims to investigate non-invasive diagnostic methods, including patient characteristics and qualitative imaging analysis as a prognostic classifier and predictor for druggable oncogenes.

METHODS

We performed a retrospective analysis on 143 GBM patients (discovery cohort). Diagnostic MRIs were re-analyzed for qualitative imaging features (VASARI features). DNA was extracted from formalin-fixed, paraffin-embedded GBM tissue of the discovery cohort for next-generation sequencing (Ion Torrent Cancer Hotspot panel v2Plus), TERT-promoter mutation and MGMT-methylation analysis. Multivariable regression analysis was used to determine the prognostic and predictive value of VASARI features.

RESULTS

Of the 143 patients, median age was 61.4 years (range 15.5–84.6) with a median overall survival of 12 months (range 0–142). We observed IDH1 R132H mutation in 8.5%, MGMT-promotor methylation in 26.1%, TERT-promotor mutation (C250T;C228T) in 69.5%, EGFR mutation in 20.3% and EGFR amplification in 37.5% of all patients. A set of eight VASARI features was identified to be associated with overall survival (p< 0.001), which is currently being validated in an external dataset (n= 184). Interestingly, VASARI features appeared to be associated with IDH1-mutation (four features, p=0.004), TERT-promotor mutation (five features, p-value < 0.001), EGFR mutation (five features, p-value < 0.001) and EGFR amplification (seven features, p-value < 0.001) but not with MGMT-methylation (two features, p-value=0.054). Additional cancer hotspots are currently being analyzed and internal validation is ongoing.

CONCLUSION AND FUTURE PERSPECTIVES

We propose an integrated prognostic classifier comprising MRI features, also associated with GBM-specific molecular alterations. Additionally, quantitative MRI radiomics features are being extracted from the discovery and validation set and incorporated in the prognostic classifier. Subsequently, radiomics and VASARI features will be correlated to intratumoral heterogeneity, assessed by tissue micro-array analysis of the discovery cohort.

Challenges and caveats of a multi-center retrospective radiomics study: an example of early treatment response assessment for NSCLC patients using FDG-PET/CT radiomics

Background

Prognostic models based on individual patient characteristics can improve treatment decisions and outcome in the future. In many (radiomic) studies, small size and heterogeneity of datasets is a challenge that often limits performance and potential clinical applicability of these models. The current study is example of a retrospective multi-centric study with challenges and caveats. To highlight common issues and emphasize potential pitfalls, we aimed for an extensive analysis of these multi-center pre-treatment datasets, with an additional ¹⁸F-fluorodeoxyglucose (FDG) positron emission tomography/computed tomography (PET/CT) scan acquired during treatment.

Methods

The dataset consisted of 138 stage II-IV non-small cell lung cancer (NSCLC) patients from four different cohorts acquired from three different institutes. The differences between the cohorts were compared in terms of clinical characteristics and using the so-called ‘cohort differences model’ approach. Moreover, the potential prognostic performances for overall survival of radiomic features extracted from CT or FDG-PET, or relative or absolute differences between the scans at the two time points, were assessed using the LASSO regression method. Furthermore, the performances of five different classifiers were evaluated for all image sets.

Results

The individual cohorts substantially differed in terms of patient characteristics. Moreover, the cohort differences model indicated statistically significant differences between the cohorts. Neither LASSO nor any of the tested classifiers resulted in a clinical relevant prognostic model that could be validated on the available datasets.

Conclusion

The results imply that the study might have been influenced by a limited sample size, heterogeneous patient characteristics, and inconsistent imaging parameters. No prognostic performance of FDG-PET or CT based radiomics models can be reported. This study highlights the necessity of extensive evaluations of cohorts and of validation datasets, especially in retrospective multi-centric datasets.

Genomics of non-small cell lung cancer (NSCLC): Association between CT-based imaging features and EGFR and K-RAS mutations in 122 patients-An external validation

OBJECTIVE

To validate previously identified associations between radiological features and clinical features with Epidermal Growth Factor Receptor (EGFR)/ Kirsten RAt Sarcoma (KRAS) alterations in an independent group of patients with Non-Small Cell Lung Cancer (NSCLC).

MATERIAL AND METHODS

A total of 122 patients with NSCLC tested for EGFR/KRAS alterations were included. Clinical and radiological features were recorded. Univariate analysis were performed to look at the associations of the studied features with EGFR/KRAS alterations. Previously calculated composite model parameters for each gene alteration prediction were applied to this validation cohort. ROC (Receiver Operating Characteristic) curves were drawn using the previously validated composite models, and also for each significant individual characteristic of the previous training cohort model. The Area Under the ROC Curve (AUC) with 95% Confidence Intervals (CI) was calculated and compared between the full models.

RESULTS

At univariate analysis, EGFR+ confirmed an association with an internal air bronchogram, pleural retraction, emphysema and lack of smoking; KRAS+ with round shape, emphysema and smoking. The AUC (95%CI) in the new cohort was confirmed to be high for EGFR+ prediction, with a value of: 0.82 (0.69-0.95) vs. 0.82 in the previous cohort, whereas it was smaller for KRAS+ prediction, with a value of 0.60 (0.48-0.72) vs. 0.67 in the previous cohort. Looking at single features in the new cohort, we found that the AUC for the models including only smoking was similar to that of the full model (including radiological and clinical features) for both gene alterations.

CONCLUSIONS

Although this study validated the significant association of clinical and radiological features with EGFR/KRAS alterations, models based on these composite features are not superior to smoking history alone to predict the mutations.

Pre-treatment CT radiomics to predict 3-year overall survival following chemoradiotherapy of esophageal cancer

BACKGROUND

Radiomic features retrieved from standard CT-images have shown prognostic power in several tumor sites. In this study, we investigated the prognostic value of pretreatment CT radiomic features to predict overall survival of esophageal cancer patients after chemoradiotherapy.

MATERIAL AND METHODS

Two datasets of independent centers were analyzed, consisting of esophageal cancer patients treated with concurrent chemotherapy (Carboplatin/Paclitaxel) and 41.4Gy radiotherapy, followed by surgery if feasible. In total, 1049 radiomic features were calculated from the primary tumor volume. Recursive feature elimination was performed to select the 40 most relevant predictors. Using these 40 features and six clinical variables as input, two random forest (RF) models predicting 3-year overall survival were developed.

RESULTS

In total 165 patients from center 1 and 74 patients from center 2 were used. The radiomics-based RF model yielded an area under the curve (AUC) of 0.69 (95%CI 0.61-0.77), with the top-5 most important features for 3-year survival describing tumor heterogeneity after wavelet filtering. In the validation dataset, the RF model yielded an AUC of 0.61 (95%CI 0.47-0.75). Kaplan Meier plots were significantly different between risk groups in the training dataset (p = .027) and borderline significant in the validation dataset (p = .053). The clinical RF model yielded AUCs of 0.63 (95%CI 0.54-0.71) and 0.62 (95%CI 0.49-0.76) in the training and validation dataset, respectively. Risk groups did not reach a significant correlation with pathological response in the primary tumor.

CONCLUSIONS

A RF model predicting 3-year overall survival based on pretreatment CT radiomic features was developed and validated in two independent datasets of esophageal cancer patients. The radiomics model had better prognostic power compared to the model using standard clinical variables.

A review on radiomics and the future of theranostics for patient selection in precision medicine

The growing complexity and volume of clinical data and the associated decision-making processes in oncology promote the advent of precision medicine. Precision (or personalised) medicine describes preventive and/or treatment procedures that take individual patient variability into account when proscribing treatment, and has been hindered in the past by the strict requirements of accurate, robust, repeatable and preferably non-invasive biomarkers to stratify both the patient and the disease. In oncology, tumour subtypes are traditionally measured through repeated invasive biopsies, which are taxing for the patient and are cost and labour intensive. Quantitative analysis of routine clinical imaging provides an opportunity to capture tumour heterogeneity non-invasively, cost-effectively and on large scale. In current clinical practice radiological images are qualitatively analysed by expert radiologists whose interpretation is known to suffer from inter- and intra-operator variability. Radiomics, the high-throughput mining of image features from medical images, provides a quantitative and robust method to assess tumour heterogeneity, and radiomics-based signatures provide a powerful tool for precision medicine in cancer treatment. This study aims to provide an overview of the current state of radiomics as a precision medicine decision support tool. We first provide an overview of the requirements and challenges radiomics currently faces in being incorporated as a tool for precision medicine, followed by an outline of radiomics' current applications in the treatment of various types of cancer. We finish with a discussion of possible future advances that can further develop radiomics as a precision medicine tool.

Development of an isotoxic decision support system integrating genetic markers of toxicity for the implantation of a rectum spacer

INTRODUCTION

Previous studies revealed that dose escalated radiotherapy for prostate cancer patients leads to higher tumor control probabilities (TCP) but also to higher rectal toxicities. An isotoxic model was developed to maximize the given dose while controlling the toxicity level. This was applied to analyze the effect of an implantable rectum spacer (IRS) and extended with a genetic test of normal tissue radio-sensitivity. A virtual IRS (V-IRS) was tested using this method. We hypothesized that the patients with increased risk of toxicity would benefit more from an IRS.

MATERIAL AND METHODS

Sixteen localized prostate cancer patients implanted with an IRS were included in the study. Treatment planning was performed on computed tomography (CT) images before and after the placement of the IRS and with a V-IRS. The normal tissue complication probability (NTCP) was calculated using a QUANTEC reviewed model for Grade > =2 late rectal bleeding and the number of fractions of the plans were adjusted until the NTCP value was under 5%. The resulting treatment plans were used to calculate the TCP before and after placement of an IRS. This was extended by adding the effect of two published genetic single nucleotide polymorphisms (SNP's) for late rectal bleeding.

RESULTS

The median TCP resulting from the optimized plans in patients before the IRS was 75.1% [32.6-90.5%]. With IRS, the median TCP is significantly higher: 98.9% [80.8-99.9%] (p < .01). The difference in TCP between the V-IRS and the real IRS was 1.8% [0.0-18.0%]. Placing an IRS in the patients with SNP's improved the TCP from 49.0% [16.1-80.8%] and 48.9% [16.0-72.8%] to 96.3% [67.0-99.5%] and 90.1% [49.0-99.5%] (p < .01) respectively for either SNP.

CONCLUSION

This study was a proof-of-concept for an isotoxic model with genetic biomarkers with a V-IRS as a multifactorial decision support system for the decision of a placement of an IRS.

A Deep Look Into the Future of Quantitative Imaging in Oncology: A Statement of Working Principles and Proposal for Change

The adoption of enterprise digital imaging, along with the development of quantitative imaging methods and the re-emergence of statistical learning, has opened the opportunity for more personalized cancer treatments through transformative data science research. In the last 5 years, accumulating evidence has indicated that noninvasive advanced imaging analytics (i.e., radiomics) can reveal key components of tumor phenotype for multiple lesions at multiple time points over the course of treatment. Many groups using homegrown software have extracted engineered and deep quantitative features on 3-dimensional medical images for better spatial and longitudinal understanding of tumor biology and for the prediction of diverse outcomes. These developments could augment patient stratification and prognostication, buttressing emerging targeted therapeutic approaches. Unfortunately, the rapid growth in popularity of this immature scientific discipline has resulted in many early publications that miss key information or use underpowered patient data sets, without production of generalizable results. Quantitative imaging research is complex, and key principles should be followed to realize its full potential. The fields of quantitative imaging and radiomics in particular require a renewed focus on optimal study design and reporting practices, standardization, interpretability, data sharing, and clinical trials. Standardization of image acquisition, feature calculation, and statistical analysis (i.e., machine learning) are required for the field to move forward. A new data-sharing paradigm enacted among open and diverse participants (medical institutions, vendors and associations) should be embraced for faster development and comprehensive clinical validation of imaging biomarkers. In this review and critique of the field, we propose working principles and fundamental changes to the current scientific approach, with the goal of high-impact research and development of actionable prediction models that will yield more meaningful applications of precision cancer medicine.

Applicability of a prognostic CT-based radiomic signature model trained on stage I-III non-small cell lung cancer in stage IV non-small cell lung cancer

OBJECTIVES

Recently it has been shown that radiomic features of computed tomography (CT) have prognostic information in stage I-III non-small cell lung cancer (NSCLC) patients. We aim to validate this prognostic radiomic signature in stage IV adenocarcinoma patients undergoing chemotherapy.

MATERIALS AND METHODS

Two datasets of chemo-naive stage IV adenocarcinoma patients were investigated, dataset 1: 285 patients with CTs performed in a single center; dataset 2: 223 patients included in a multicenter clinical trial. The main exclusion criteria were EGFR mutation or unknown mutation status and non-delineated primary tumor. Radiomic features were calculated for the primary tumor. The c-index of cox regression was calculated and compared to the signature performance for overall survival (OS).

RESULTS

In total CT scans from 195 patients were eligible for analysis. Patients having a prognostic index (PI) lower than the signature median (n = 92) had a significantly better OS than patients with a PI higher than the median (n = 103, HR 1.445, 95% CI 1.07-1.95, p = 0.02, c-index 0.576, 95% CI 0.527-0.624).

CONCLUSION

The radiomic signature, derived from daily practice CT scans, has prognostic value for stage IV NSCLC, however the signature performs less than previously described for stage I-III NSCLC stages. In the future, machine learning techniques can potentially lead to a better prognostic imaging based model for stage IV NSCLC.

Machine learning algorithms for outcome prediction in (chemo)radiotherapy: An empirical comparison of classifiers

Purpose:

Machine learning classification algorithms (classifiers) for prediction of treatment response are becoming more popular in radiotherapy literature. General Machine learning literature provides evidence in favor of some classifier families (random forest, support vector machine, gradient boosting) in terms of classification performance. The purpose of this study is to compare such classifiers specifically for (chemo)radiotherapy datasets and to estimate their average discriminative performance for radiation treatment outcome prediction.

Methods:

We collected 12 datasets (3496 patients) from prior studies on post-(chemo)radiotherapy toxicity, survival, or tumor control with clinical, dosimetric, or blood biomarker features from multiple institutions and for different tumor sites, that is, (non-)small-cell lung cancer, head and neck cancer, and meningioma. Six common classification algorithms with built-in feature selection (decision tree, random forest, neural network, support vector machine, elastic net logistic regression, Logit-Boost) were applied on each dataset using the popular open-source R package caret. The R code and documentation for the analysis are available online (https://github.com/timodeist/classifier_selection_code). All classifiers were run on each dataset in a 100-repeated nested fivefold cross-validation with hyperparameter tuning. Performance metrics (AUC, calibration slope and intercept, accuracy, Cohen’s kappa, and Brier score) were computed. We ranked classifiers by AUC to determine which classifier is likely to also perform well in future studies. We simulated the benefit for potential investigators to select a certain classifier for a new dataset based on our study (pre-selection based on other datasets) or estimating the best classifier for a dataset (set-specific selection based on information from the new dataset) compared with uninformed classifier selection (random selection).

Results:

Random forest (best in 6/12 datasets) and elastic net logistic regression (best in 4/12 datasets) showed the overall best discrimination, but there was no single best classifier across datasets. Both classifiers had a median AUC rank of 2. Preselection and set-specific selection yielded a significant average AUC improvement of 0.02 and 0.02 over random selection with an average AUC rank improvement of 0.42 and 0.66, respectively.

Conclusion:

Random forest and elastic net logistic regression yield higher discriminative performance in (chemo)radiotherapy outcome and toxicity prediction than other studied classifiers. Thus, one of these two classifiers should be the first choice for investigators when building classification models or to benchmark one’s own modeling results against. Our results also show that an informed preselection of classifiers based on existing datasets can improve discrimination over random selection.

Development and validation of a radiomic signature to predict HPV (p16) status from standard CT imaging: a multicenter study

OBJECTIVES

Human papillomavirus (HPV) positive oropharyngeal cancer (oropharyngeal squamous cell carcinoma, OPSCC) is biologically and clinically different from HPV negative OPSCC. Here, we evaluate the use of a radiomic approach to identify the HPV status of OPSCC.

METHODS

Four independent cohorts, totaling 778 OPSCC patients with HPV determined by p16 were collected. We randomly assigned 80% of all data for model training (N = 628) and 20% for validation (N = 150). On the pre-treatment CT images, 902 radiomic features were calculated from the gross tumor volume. Multivariable modeling was performed using least absolute shrinkage and selection operator. To assess the impact of CT artifacts in predicting HPV (p16), a model was developed on all training data (Mall) and on the artifact-free subset of training data (Mno art). Models were validated on all validation data (Vall), and the subgroups with (Vart) and without (Vno art) artifacts. Kaplan-Meier survival analysis was performed to compare HPV status based on p16 and radiomic model predictions.

RESULTS

The area under the receiver operator curve for Mall and Mno art ranged between 0.70 and 0.80 and was not significantly different for all validation data sets. There was a consistent and significant split between survival curves with HPV status determined by p16 [p = 0.007; hazard ratio (HR): 0.46], Mall (p = 0.036; HR: 0.55) and Mno art (p = 0.027; HR: 0.49).

CONCLUSION

This study provides proof of concept that molecular information can be derived from standard medical images and shows potential for radiomics as imaging biomarker of HPV status. Advances in knowledge: Radiomics has the potential to identify clinically relevant molecular phenotypes.

Stereotactic Radiosurgery in the Management of Patients With Brain Metastases of Non-Small Cell Lung Cancer: Indications, Decision Tools and Future Directions

Brain metastases (BM) frequently occur in non-small cell lung cancer (NSCLC) patients. Most patients with BM have a limited life expectancy, measured in months. Selected patients may experience a very long progression-free survival, for example, patients with a targetable driver mutation. Traditionally, whole-brain radiotherapy (WBRT) has been the cornerstone of the treatment, but its indication is a matter of debate. A randomized trial has shown that for patients with a poor prognosis, WBRT does not add quality of life (QoL) nor survival over the best supportive care. In recent decades, stereotactic radiosurgery (SRS) has become an attractive non-invasive treatment for patients with BM. Only the BM is irradiated to an ablative dose, sparing healthy brain tissue. Intracranial recurrence rates decrease when WBRT is administered following SRS or resection but does not improve overall survival and comes at the expense of neurocognitive function and QoL. The downside of SRS compared with WBRT is a risk of radionecrosis (RN) and a higher risk of developing new BM during follow-up. Currently, SRS is an established treatment for patients with a maximum of four BM. Several promising strategies are currently being investigated to further improve the indication and outcome of SRS for patients with BM: the effectivity and safety of SRS in patients with more than four BM, combining SRS with systemic therapy such as targeted agents or immunotherapy, shared decision-making with SRS as a treatment option, and individualized isotoxic dose prescription to mitigate the risk of RN and further enhance local control probability of SRS. This review discusses the current indications of SRS and future directions of treatment for patients with BM of NSCLC with focus on the value of SRS.

Results and adverse events of personalized peptide receptor radionuclide therapy with 90Yttrium and 177Lutetium in 1048 patients with neuroendocrine neoplasms

Introduction

Peptide receptor radionuclide therapy (PRRT) of patients with somatostatin receptor expressing neuroendocrine neoplasms has shown promising results in clinical trials and a recently published phase III study.

Methods

In our center, 2294 patients were screened between 2004 and 2014 by ⁶⁸Ga somatostatin receptor (SSTR) PET/CT. Intention to treat analysis included 1048 patients, who received at least one cycle of ⁹⁰Yttrium or ¹⁷⁷Lutetium-based PRRT. Progression free survival was determined by ⁶⁸Ga SSTR-PET/CT and EORTC response criteria. Adverse events were determined by CTCAE criteria.

Results

Overall survival (95% confidence interval) of all patients was 51 months (47.0-54.9) and differed significantly according to radionuclide, grading, previous therapies, primary site and functionality. Progression free survival (based on PET/CT) of all patients was 19 months (16.9-21), which was significantly influenced by radionuclide, grading, and origin of neuroendocrine neoplasm. Progression free survival after initial progression and first and second resumption of PRRT after therapy-free intervals of more than 6 months were 11 months (9.4-12.5) and 8 months (6.4-9.5), respectively. Myelodysplastic syndrome or leukemia developed in 22 patients (2.1%) and 5 patients required hemodialysis after treatment, other adverse events were rare.

Conclusion

PRRT is effective and overall survival is favorable in patients with neuroendocrine neoplasms depending on the radionuclide used for therapy, grading and origin of the neuroendocrine neoplasm which is not exactly mirrored in progression free survival as determined by highly sensitive ⁶⁸Ga somatostatin receptor PET/CT using EORTC criteria for determining response to therapy.

A prediction model for early death in non-small cell lung cancer patients following curative-intent chemoradiotherapy

BACKGROUND

Early death after a treatment can be seen as a therapeutic failure. Accurate prediction of patients at risk for early mortality is crucial to avoid unnecessary harm and reducing costs. The goal of our work is two-fold: first, to evaluate the performance of a previously published model for early death in our cohorts. Second, to develop a prognostic model for early death prediction following radiotherapy.

MATERIAL AND METHODS

Patients with NSCLC treated with chemoradiotherapy or radiotherapy alone were included in this study. Four different cohorts from different countries were available for this work (N = 1540). The previous model used age, gender, performance status, tumor stage, income deprivation, no previous treatment given (yes/no) and body mass index to make predictions. A random forest model was developed by learning on the Maastro cohort (N = 698). The new model used performance status, age, gender, T and N stage, total tumor volume (cc), total tumor dose (Gy) and chemotherapy timing (none, sequential, concurrent) to make predictions. Death within 4 months of receiving the first radiotherapy fraction was used as the outcome.

RESULTS

Early death rates ranged from 6 to 11% within the four cohorts. The previous model performed with AUC values ranging from 0.54 to 0.64 on the validation cohorts. Our newly developed model had improved AUC values ranging from 0.62 to 0.71 on the validation cohorts.

CONCLUSIONS

Using advanced machine learning methods and informative variables, prognostic models for early mortality can be developed. Development of accurate prognostic tools for early mortality is important to inform patients about treatment options and optimize care.

Predicting tumor hypoxia in non-small cell lung cancer by combining CT, FDG PET and dynamic contrast-enhanced CT

BACKGROUND

Most solid tumors contain inadequately oxygenated (i.e., hypoxic) regions, which tend to be more aggressive and treatment resistant. Hypoxia PET allows visualization of hypoxia and may enable treatment adaptation. However, hypoxia PET imaging is expensive, time-consuming and not widely available. We aimed to predict hypoxia levels in non-small cell lung cancer (NSCLC) using more easily available imaging modalities: FDG-PET/CT and dynamic contrast-enhanced CT (DCE-CT).

MATERIAL AND METHODS

For 34 NSCLC patients, included in two clinical trials, hypoxia HX4-PET/CT, planning FDG-PET/CT and DCE-CT scans were acquired before radiotherapy. Scans were non-rigidly registered to the planning CT. Tumor blood flow (BF) and blood volume (BV) were calculated by kinetic analysis of DCE-CT images. Within the gross tumor volume, independent clusters, i.e., supervoxels, were created based on FDG-PET/CT. For each supervoxel, tumor-to-background ratios (TBR) were calculated (median SUV/aorta SUV_mean) for HX4-PET/CT and supervoxel features (median, SD, entropy) for the other modalities. Two random forest models (cross-validated: 10 folds, five repeats) were trained to predict the hypoxia TBR; one based on CT, FDG, BF and BV, and one with only CT and FDG features. Patients were split in a training (trial NCT01024829) and independent test set (trial NCT01210378). For each patient, predicted, and observed hypoxic volumes (HV) (TBR > 1.2) were compared.

RESULTS

Fifteen patients (3291 supervoxels) were used for training and 19 patients (1502 supervoxels) for testing. The model with all features (RMSE training: 0.19 ± 0.01, test: 0.27) outperformed the model with only CT and FDG-PET features (RMSE training: 0.20 ± 0.01, test: 0.29). All tumors of the test set were correctly classified as normoxic or hypoxic (HV > 1 cm³) by the best performing model.

CONCLUSIONS

We created a data-driven methodology to predict hypoxia levels and hypoxia spatial patterns using CT, FDG-PET and DCE-CT features in NSCLC. The model correctly classifies all tumors, and could therefore, aid tumor hypoxia classification and patient stratification.

Developing and Validating a Survival Prediction Model for NSCLC Patients Through Distributed Learning Across 3 Countries

PURPOSE

Tools for survival prediction for non-small cell lung cancer (NSCLC) patients treated with chemoradiation or radiation therapy are of limited quality. In this work, we developed a predictive model of survival at 2 years. The model is based on a large volume of historical patient data and serves as a proof of concept to demonstrate the distributed learning approach.

METHODS AND MATERIALS

Clinical data from 698 lung cancer patients, treated with curative intent with chemoradiation or radiation therapy alone, were collected and stored at 2 different cancer institutes (559 patients at Maastro clinic (Netherlands) and 139 at Michigan university [United States]). The model was further validated on 196 patients originating from The Christie (United Kingdon). A Bayesian network model was adapted for distributed learning (the animation can be viewed at https://www.youtube.com/watch?v=ZDJFOxpwqEA). Two-year posttreatment survival was chosen as the endpoint. The Maastro clinic cohort data are publicly available at https://www.cancerdata.org/publication/developing-and-validating-survival-prediction-model-nsclc-patients-through-distributed, and the developed models can be found at www.predictcancer.org.

RESULTS

Variables included in the final model were T and N category, age, performance status, and total tumor dose. The model has an area under the curve (AUC) of 0.66 on the external validation set and an AUC of 0.62 on a 5-fold cross validation. A model based on the T and N category performed with an AUC of 0.47 on the validation set, significantly worse than our model (P<.001). Learning the model in a centralized or distributed fashion yields a minor difference on the probabilities of the conditional probability tables (0.6%); the discriminative performance of the models on the validation set is similar (P=.26).

CONCLUSIONS

Distributed learning from federated databases allows learning of predictive models on data originating from multiple institutions while avoiding many of the data-sharing barriers. We believe that distributed learning is the future of sharing data in health care.

Infrastructure and distributed learning methodology for privacy-preserving multi-centric rapid learning health care: euroCAT

•

Developed and implemented IT infrastructure in 5 radiation clinics across 3 countries.

•

Proof-of-principle for ‘big data’ infrastructure and distributed learning studies.

•

General framework to execute learning algorithms on distributed data.

Machine learning applications for personalized medicine are highly dependent on access to sufficient data. For personalized radiation oncology, datasets representing the variation in the entire cancer patient population need to be acquired and used to learn prediction models. Ethical and legal boundaries to ensure data privacy hamper collaboration between research institutes. We hypothesize that data sharing is possible without identifiable patient data leaving the radiation clinics and that building machine learning applications on distributed datasets is feasible.

We developed and implemented an IT infrastructure in five radiation clinics across three countries (Belgium, Germany, and The Netherlands). We present here a proof-of-principle for future ‘big data’ infrastructures and distributed learning studies. Lung cancer patient data was collected in all five locations and stored in local databases. Exemplary support vector machine (SVM) models were learned using the Alternating Direction Method of Multipliers (ADMM) from the distributed databases to predict post-radiotherapy dyspnea grade ⩾2. The discriminative performance was assessed by the area under the curve (AUC) in a five-fold cross-validation (learning on four sites and validating on the fifth). The performance of the distributed learning algorithm was compared to centralized learning where datasets of all institutes are jointly analyzed.

The euroCAT infrastructure has been successfully implemented in five radiation clinics across three countries. SVM models can be learned on data distributed over all five clinics. Furthermore, the infrastructure provides a general framework to execute learning algorithms on distributed data. The ongoing expansion of the euroCAT network will facilitate machine learning in radiation oncology. The resulting access to larger datasets with sufficient variation will pave the way for generalizable prediction models and personalized medicine.

Individualized early death and long-term survival prediction after stereotactic radiosurgery for brain metastases of non-small cell lung cancer: Two externally validated nomograms

INTRODUCTION

Commonly used clinical models for survival prediction after stereotactic radiosurgery (SRS) for brain metastases (BMs) are limited by the lack of individual risk scores and disproportionate prognostic groups. In this study, two nomograms were developed to overcome these limitations.

METHODS

495 patients with BMs of NSCLC treated with SRS for a limited number of BMs in four Dutch radiation oncology centers were identified and divided in a training cohort (n=214, patients treated in one hospital) and an external validation cohort n=281, patients treated in three other hospitals). Using the training cohort, nomograms were developed for prediction of early death (<3months) and long-term survival (>12months) with prognostic factors for survival. Accuracy of prediction was defined as the area under the curve (AUC) by receiver operating characteristics analysis for prediction of early death and long term survival. The accuracy of the nomograms was also tested in the external validation cohort.

RESULTS

Prognostic factors for survival were: WHO performance status, presence of extracranial metastases, age, GTV largest BM, and gender. Number of brain metastases and primary tumor control were not prognostic factors for survival. In the external validation cohort, the nomogram predicted early death statistically significantly better (p<0.05) than the unfavorable groups of the RPA, DS-GPA, GGS, SIR, and Rades 2015 (AUC=0.70 versus range AUCs=0.51-0.60 respectively). With an AUC of 0.67, the other nomogram predicted 1year survival statistically significantly better (p<0.05) than the favorable groups of four models (range AUCs=0.57-0.61), except for the SIR (AUC=0.64, p=0.34). The models are available on www.predictcancer.org.

CONCLUSION

The nomograms predicted early death and long-term survival more accurately than commonly used prognostic scores after SRS for a limited number of BMs of NSCLC. Moreover these nomograms enable individualized probability assessment and are easy into use in routine clinical practice.

Decision support systems for personalized and participative radiation oncology

A paradigm shift from current population based medicine to personalized and participative medicine is underway. This transition is being supported by the development of clinical decision support systems based on prediction models of treatment outcome. In radiation oncology, these models 'learn' using advanced and innovative information technologies (ideally in a distributed fashion - please watch the animation: http://youtu.be/ZDJFOxpwqEA) from all available/appropriate medical data (clinical, treatment, imaging, biological/genetic, etc.) to achieve the highest possible accuracy with respect to prediction of tumor response and normal tissue toxicity. In this position paper, we deliver an overview of the factors that are associated with outcome in radiation oncology and discuss the methodology behind the development of accurate prediction models, which is a multi-faceted process. Subsequent to initial development/validation and clinical introduction, decision support systems should be constantly re-evaluated (through quality assurance procedures) in different patient datasets in order to refine and re-optimize the models, ensuring the continuous utility of the models. In the reasonably near future, decision support systems will be fully integrated within the clinic, with data and knowledge being shared in a standardized, dynamic, and potentially global manner enabling truly personalized and participative medicine.

Big Data in radiation therapy: challenges and opportunities

Data collected and generated by radiation oncology can be classified by the Volume, Variety, Velocity and Veracity (4Vs) of Big Data because they are spread across different care providers and not easily shared owing to patient privacy protection. The magnitude of the 4Vs is substantial in oncology, especially owing to imaging modalities and unclear data definitions. To create useful models ideally all data of all care providers are understood and learned from; however, this presents challenges in the guise of poor data quality, patient privacy concerns, geographical spread, interoperability and large volume. In radiation oncology, there are many efforts to collect data for research and innovation purposes. Clinical trials are the gold standard when proving any hypothesis that directly affects the patient. Collecting data in registries with strict predefined rules is also a common approach to find answers. A third approach is to develop data stores that can be used by modern machine learning techniques to provide new insights or answer hypotheses. We believe all three approaches have their strengths and weaknesses, but they should all strive to create Findable, Accessible, Interoperable, Reusable (FAIR) data. To learn from these data, we need distributed learning techniques, sending machine learning algorithms to FAIR data stores around the world, learning from trial data, registries and routine clinical data rather than trying to centralize all data. To improve and personalize medicine, rapid learning platforms must be able to process FAIR “Big Data” to evaluate current clinical practice and to guide further innovation.