Machine Learning Algorithm to Estimate Distant Breast Cancer Recurrence at the Population Level with Administrative Data.

Purpose: High-quality population-based cancer recurrence data are scarcely available, mainly due to complexity and cost of registration. For the first time in Belgium, we developed a tool to estimate distant recurrence after a breast cancer diagnosis at the population level, based on real-world cancer registration and administrative data. Methods: Data on distant cancer recurrence (including progression) from patients diagnosed with breast cancer between 2009-2014 were collected from medical files at 9 Belgian centers to train, test and externally validate an algorithm (i.e., gold standard). Distant recurrence was defined as the occurrence of distant metastases between 120 days and within 10 years after the primary diagnosis, with follow-up until December 31, 2018. Data from the gold standard were linked to population-based data from the Belgian Cancer Registry (BCR) and administrative data sources. Potential features to detect recurrences in administrative data were defined based on expert opinion from breast oncologists, and subsequently selected using bootstrap aggregation. Based on the selected features, classification and regression tree (CART) analysis was performed to construct an algorithm for classifying patients as having a distant recurrence or not. Results: A total of 2507 patients were included of whom 216 had a distant recurrence in the clinical data set. The performance of the algorithm showed sensitivity of 79.5% (95% CI 68.8-87.8%), positive predictive value (PPV) of 79.5% (95% CI 68.8-87.8%), and accuracy of 96.7% (95% CI 95.4-97.7%). The external validation resulted in a sensitivity of 84.1% (95% CI 74.4-91.3%), PPV of 84.1% (95% CI 74.4-91.3%), and an accuracy of 96.8% (95% CI 95.4-97.9%). Conclusion: Our algorithm detected distant breast cancer recurrences with an overall good accuracy of 96.8% for patients with breast cancer, as observed in the first multi-centric external validation exercise.

The epidemiology of multiple primary cancers in Belgium (2004-2017): Incidence, proportion, risk, stage and impact on relative survival estimates.

BACKGROUND: As both life expectancy and cancer survival improve, the incidence of multiple primary cancer has augmented and is expected to further increase. This study describes for the first time the epidemiology of multiple invasive tumours in Belgium. METHODS: This nationwide study, based on all cancers diagnosed between 2004 and 2017 in Belgium, describes the proportion of multiple primary cancer, its evolution over time, the impact of inclusion or exclusion of multiple primary cancer on relative survival estimates, the risk of developing a second primary cancer, and the difference in stage between first and second primary cancer for the same patient. RESULTS: The proportion of multiple primary cancer increases with age, varies across cancer sites (from 4% for testis cancer to 22.8% for oesophageal cancer), is higher in men than in women, and has linearly increased over time. The inclusion of multiple primary cancer resulted in smaller 5-year relative survival and this impact is more pronounced in cancer sites with high relative survival. Patients with a first primary cancer have an increased risk to develop a new primary cancer compared to the population without a previous cancer history (1.27 and 1.59 times higher in men and women, respectively) and this risk depends on cancer site. Second primary cancers are associated with more advanced stages and more unknown stages than the corresponding first cancer diagnosis. CONCLUSIONS: This study describes multiple primary cancer according to several measures (proportion, standardised incidence ratio for an second primary cancer, impact of multiple primary cancer on relative survival and differences according to stage) for the first time in Belgium. The results are based on data of a population-based cancer registry with a relatively recent onset (2004).

Real-World Estimation of First- and Second-Line Treatments for Diffuse Large B-Cell Lymphoma Using Health Insurance Data: A Belgian Population-Based Study.

We determined first- and second-line regimens, including hematopoietic stem cell transplantations, in all diffuse large B cell lymphoma (DLBCL) patients aged ≥20 yr (n = 1,888), registered at the Belgian Cancer Registry (2013-2015). Treatments were inferred from reimbursed drugs, and procedures registered in national health insurance databases. This real-world population-based study allows to assess patients usually excluded from clinical trials such as those with comorbidities, other malignancies (12%), and advanced age (28% are ≥80 yr old). Our data show that the majority of older patients are still started on first-line regimens with curative intent and a substantial proportion of them benefit from this approach. First-line treatments included full R-CHOP (44%), "incomplete" (R-)CHOP (18%), other anthracycline (14%), non-anthracycline (9%), only radiotherapy (3%), and no chemo-/radiotherapy (13%), with significant variation between age groups. The 5-year overall survival (OS) of all patients was 56% with a clear influence of age (78% [20-59 yr] versus 16% [≥85 yr]) and of the type of first-line treatments: full R-CHOP (72%), other anthracycline (58%), "incomplete" (R-)CHOP (47%), non-anthracycline (30%), only radiotherapy (30%), and no chemo-/radiotherapy (9%). Second-line therapy, presumed for refractory (7%) or relapsed disease (9%), was initiated in 252 patients (16%) and was predominantly (71%) platinum-based. The 5-year OS after second-line treatment without autologous stem cell transplantation (ASCT) was generally poor (11% in ≥70 yr versus 17% in <70 yr). An ASCT was performed in 5% of treated patients (n = 82). The 5-year OS after first- or second-line ASCT was similar (69% versus 66%). After adjustment, multivariable OS analyses indicated a significant hazard ratio (HR) for, among others, age (HR 1.81 to 5.95 for increasing age), performance status (PS) (HR 4.56 for PS >1 within 3 months from incidence), subsequent malignancies (HR 2.50), prior malignancies (HR 1.34), respiratory and diabetic comorbidity (HR 1.41 and 1.24), gender (HR 1.25 for males), and first-line treatment with full R-CHOP (HR 0.41) or other anthracycline-containing regimens (HR 0.72). Despite inherent limitations, patterns of care in DLBCL could be determined using an innovative approach based on Belgian health insurance data.

The Impact of Regional Variation in Clinical Practice on Thyroid Cancer Diagnosis: A National Population-Based Study.

OBJECTIVE: Regional variation in thyroid cancer incidence in Belgium, most pronounced for low risk cancer, was previously shown to be related to variation in clinical practice, with higher thyroid surgery rates and lower proportions of preoperative fine-needle aspiration (FNA) in regions with high thyroid cancer incidence (period 2004-2006). The objective of this study was to investigate regional thyroid cancer incidence variation in relation with variation in thyroid surgery threshold in a more recent Belgian thyroid cancer cohort. METHODS: A population-based cohort of thyroid cancer patients that underwent a (near) total thyroidectomy in the period 2009-2011 (n = 2,329 patients) was identified and studied by linking data from the Belgian cancer registry and the Belgian health insurance companies, and case-by-case study of the pathology protocols. The execution of preoperative FNA and the thyroid resection specimen weight were compared between high and low thyroid cancer incidence regions. Thyroid weight in the pT1a-restricted group was studied as a proxy for surgical threshold for benign nodular goiter. Furthermore, time trend analyses were performed for the execution of FNA for the period 2004-2012. RESULTS: Although a lower proportion of FNA in the high thyroid cancer incidence region persisted in the period 2009-2011 (41.2% [31.9-50.9] vs. 72.9% [64.9-79.7] in the low-incidence region (LIR), p < 0.001), a positive time trend was observed for the period 2004-2012. The median thyroid surgical specimen weight was lower in the high incidence region compared to the LIR (27.0 g [IQR 18.0-45.3] vs. 36.0 g [IQR 22.0-73.0], p < 0.0001), and this finding was corroborated in the pT1a-restricted group. CONCLUSION: Interregional differences in use of FNA and surgical thyroid specimen weight are consistent with an inverse relation between thyroid cancer incidence and thyroid surgery threshold, carrying risk for overdiagnosis.

Comorbidity in head and neck cancer: Is it associated with therapeutic delay, post-treatment mortality and survival in a population-based study?

OBJECTIVES: This study aims to investigate the relationship between comorbidities and therapeutic delay, post-treatment mortality, overall and relative survival in patients diagnosed with squamous cell carcinoma of the head and neck (HNSCC). PATIENTS AND METHODS: 9245 patients with a single HNSCC diagnosed between 2009 and 2014 were identified in the Belgian Cancer Registry. The Charlson Comorbidity Index (CCI) was calculated for 8812 patients (95.3%), distinguishing patients having none (0), mild (1-2), moderate (3-4) or severe comorbidity (>4). The relationship between CCI and therapeutic delay was evaluated using the Spearman correlation. Post-treatment mortality was modelled with logistic regression, using death within 30 days as the event. The association between comorbidity and survival was assessed using Cox proportional hazard models. RESULTS: Among 8812 patients with a known CCI, 39.2% had at least one comorbidity. Therapeutic delay increased from 31 to 36 days when the CCI worsened from 0 to 4 (rho = 0.087). After case-mix adjustment, higher baseline comorbidity was associated with increased post-surgery mortality (mild, OR 3.52 [95% CI 1.91-6.49]; severe, OR 18.71 [95% CI 6.85-51.12]) and post-radiotherapy mortality (mild, OR 2.23 [95% CI 1.56-3.19]; severe, OR 9.33 [95% CI 4.83-18.01]) and with reduced overall survival (mild, HR 1.39, [95% CI 1.31-1.48]; severe, HR 2.41 [95% CI 2.00-2.90]). That was also the case for relative survival in unadjusted analyses (mild, EHR 1.77 [95% CI 1.64-1.92]; severe, EHR = 4.15 [95% CI 3.43-5.02]). CONCLUSION: Comorbidity is significantly related to therapeutic delay, post-treatment mortality, 5-year overall and relative survival in HNSCC patients. Therapeutic decision support tools should optimally integrate comorbidity.

Benchmarking of survival outcomes following haematopoietic stem cell transplantation: A review of existing processes and the introduction of an international system from the European Society for Blood and Marrow Transplantation (EBMT) and the Joint Accreditation Committee of ISCT and EBMT (JACIE).

In many healthcare settings, benchmarking for complex procedures has become a mandatory requirement by competent authorities, regulators, payers and patients to assure clinical performance, cost-effectiveness and safe care of patients. In several countries inside and outside Europe, benchmarking systems have been established for haematopoietic stem cell transplantation (HSCT), but access is not universal. As benchmarking is now integrated into the FACT-JACIE standards, the EBMT and JACIE established a Clinical Outcomes Group (COG) to develop and introduce a universal system accessible across EBMT members. Established systems from seven European countries (United Kingdom, Italy, Belgium, France, Germany, Spain, Switzerland), USA and Australia were appraised, revealing similarities in process, but wide variations in selection criteria and statistical methods. In tandem, the COG developed the first phase of a bespoke risk-adapted international benchmarking model for one-year survival following allogeneic and autologous HSCT based on current capabilities within the EBMT registry core dataset. Data completeness, which has a critical impact on validity of centre comparisons, is also assessed. Ongoing development will include further scientific validation of the model, incorporation of further variables (when appropriate) alongside implementation of systems for clinically meaningful interpretation and governance aiming to maximise acceptance to centres, clinicians, payers and patients across EBMT.

Correction: Benchmarking of survival outcomes following haematopoietic stem cell transplantation: A review of existing processes and the introduction of an international system from the European Society for Blood and Marrow Transplantation (EBMT) and the Joint Accreditation Committee of ISCT and EBMT (JACIE).

An amendment to this paper has been published and can be accessed via a link at the top of the paper.

Providing both autologous and allogeneic hematopoietic stem cell transplants (HSCT) may have a stronger impact on the outcome of autologous HSCT in adult patients than activity levels or implementation of JACIE at Belgian transplant centres.

While performance since the introduction of the JACIE quality management system has been shown to be improved for allogeneic hematopoietic stem cell transplants (HSCT), impact on autologous-HSCT remains unclear in Europe. Our study on 2697 autologous-HSCT performed in adults in 17 Belgian centres (2007-2013) aims at comparing the adjusted 1 and 3-yr survival between the different centres & investigating the impact of 3 centre-related factors on performance (time between JACIE accreditation achievement by the centre and the considered transplant, centre activity volume and type of HSCT performed by centres: exclusively autologous vs both autologous & allogeneic). We showed a relatively homogeneous performance between Belgian centres before national completeness of JACIE implementation. The 3 centre-related factors had a significant impact on the 1-yr survival, while activity volume and type of HSCT impacted the 3-yr survival of autologous-HSCT patients in univariable analyses. Only activity volume (impact on 1-yr survival only) and type of HSCT (impact on 1 and 3-yr survivals) remained significant in multivariable analysis. This is explained by the strong relationship between these 3 variables. An extended transplantation experience, i.e., performing both auto & allo-HSCT, appears to be a newly informative quality indicator potentially conveying a multitude of underlying complex factors.

Real-life treatment practice for malignant pleural mesothelioma in Belgium.

OBJECTIVES: Malignant pleural mesothelioma (MPM) is a rare and aggressive cancer, for which treatment is often limited to palliative combination chemotherapy. Multimodality-therapy, including radical surgery, is largely restricted to clinical trials, leaving its benefit currently unclear. This study aimed to get a comprehensive view on real-world MPM treatment at the Belgian population level, to assess survival and to identify prognostic factors. MATERIALS AND METHODS: The study period covered the incidence years 2004-2012 (N = 1453). Starting from the Belgian Cancer Registry, additional information regarding patient characteristics, diagnosis and treatment was retrieved from multiple data sources. Adjusted cox proportional-hazard regression models using time-dependent covariates were performed to assess survival in relation to treatment patterns and centre volume. RESULTS: Sixty-nine percent of patients underwent tumour-directed treatment, mostly cisplatin-pemetrexed chemotherapy. Radical surgery was mainly performed in younger patients with epithelioid subtype. Centre volume, surgery and chemotherapy showed a positive relation with survival in univariable analyses, but only chemotherapy remained significantly relevant in multivariable analyses. Younger patients, females, and epithelioid subtypes also independently had a better survival. CONCLUSION: This large population-based study provides insights in MPM treatment practice in Belgium. Centre volume and surgery being related to survival in univariable analyses, only chemotherapy remained prognostic after adjustment.

Do refined consensus guidelines improve the uniformity of clinical target volume delineation for rectal cancer? Results of a national review project.

In a previous national central review project, 74% of the rectal cancer clinical target volumes (CTVs) needed a modification. In a follow-up initiative, we evaluated whether the use of refined international consensus guidelines improves the uniformity of CTV delineation in clinical practice.