Clinical bioinformatics desiderata for molecular tumor boards.

Clinical Bioinformatics is a knowledge framework required to interpret data of medical interest via computational methods. This area became of dramatic importance in precision oncology, fueled by cancer genomic profiling: most definitions of Molecular Tumor Boards require the presence of bioinformaticians. However, all available literature remained rather vague on what are the specific needs in terms of digital tools and expertise to tackle and interpret genomics data to assign novel targeted or biomarker-driven targeted therapies to cancer patients. To fill this gap, in this article, we present a catalog of software families and human skills required for the tumor board bioinformatician, with specific examples of real-world applications associated with each element presented.

Comprehensive genomic profiling on metastatic Melanoma: results from a network screening from 7 Italian Cancer Centres.

BACKGROUND: The current therapeutic algorithm for Advanced Stage Melanoma comprises of alternating lines of Targeted and Immuno-therapy, mostly via Immune-Checkpoint blockade. While Comprehensive Genomic Profiling of solid tumours has been approved as a companion diagnostic, still no approved predictive biomarkers are available for Melanoma aside from BRAF mutations and the controversial Tumor Mutational Burden. This study presents the results of a Multi-Centre Observational Clinical Trial of Comprehensive Genomic Profiling on Target and Immuno-therapy treated advanced Melanoma. METHODS: 82 samples, collected from 7 Italian Cancer Centres of FFPE-archived Metastatic Melanoma and matched blood were sequenced via a custom-made 184-gene amplicon-based NGS panel. Sequencing and bioinformatics analysis was performed at a central hub. Primary analysis was carried out via the Ion Reporter framework. Secondary analysis and Machine Learning modelling comprising of uni and multivariate, COX/Lasso combination, and Random Forest, was implemented via custom R/Python scripting. RESULTS: The genomics landscape of the ACC-mela cohort is comparable at the somatic level for Single Nucleotide Variants and INDELs aside a few gene targets. All the clinically relevant targets such as BRAF and NRAS have a comparable distribution thus suggesting the value of larger scale sequencing in melanoma. No comparability is reached at the CNV level due to biotechnological biases and cohort numerosity. Tumour Mutational Burden is slightly higher in median for Complete Responders but fails to achieve statistical significance in Kaplan-Meier survival analysis via several thresholding strategies. Mutations on PDGFRB, NOTCH3 and RET were shown to have a positive effect on Immune-checkpoint treatment Overall and Disease-Free Survival, while variants in NOTCH4 were found to be detrimental for both endpoints. CONCLUSIONS: The results presented in this study show the value and the challenge of a genomics-driven network trial. The data can be also a valuable resource as a validation cohort for Immunotherapy and Target therapy genomic biomarker research.

It is time to implement molecular classification in endometrial cancer.

A huge effort has been done in redefining endometrial cancer (EC) risk classes in the last decade. However, known prognostic factors (FIGO staging and grading, biomolecular classification and ESMO-ESGO-ESTRO risk classes stratification) are not able to predict outcomes and especially recurrences. Biomolecular classification has helped in re-classifying patients for a more appropriate adjuvant treatment and clinical studies suggest that currently used molecular classification improves the risk assessment of women with EC, however, it does not clearly explain differences in recurrence profiles. Furthermore, a lack of evidence appears in EC guidelines. Here, we summarize the main concepts why molecular classification is not enough in the management of endometrial cancer, by highlighting some promising innovative examples in scientific literature studies with a clinical potential significant impact.

The Molecular Tumor Board of the Regina Elena National Cancer Institute: from accrual to treatment in real-world.

BACKGROUND: Molecular Tumor Boards (MTB) operating in real-world have generated limited consensus on good practices for accrual, actionable alteration mapping, and outcome metrics. These topics are addressed herein in 124 MTB patients, all real-world accrued at progression, and lacking approved therapy options. METHODS: Actionable genomic alterations identified by tumor DNA (tDNA) and circulating tumor DNA (ctDNA) profiling were mapped by customized OncoKB criteria to reflect diagnostic/therapeutic indications as approved in Europe. Alterations were considered non-SoC when mapped at either OncoKB level 3, regardless of tDNA/ctDNA origin, or at OncoKB levels 1/2, provided they were undetectable in matched tDNA, and had not been exploited in previous therapy lines. RESULTS: Altogether, actionable alterations were detected in 54/124 (43.5%) MTB patients, but only in 39 cases (31%) were these alterations (25 from tDNA, 14 from ctDNA) actionable/unexploited, e.g. they had not resulted in the assignment of pre-MTB treatments. Interestingly, actionable and actionable/unexploited alterations both decreased (37.5% and 22.7% respectively) in a subset of 88 MTB patients profiled by tDNA-only, but increased considerably (77.7% and 66.7%) in 18 distinct patients undergoing combined tDNA/ctDNA testing, approaching the potential treatment opportunities (76.9%) in 147 treatment-naïve patients undergoing routine tDNA profiling for the first time. Non-SoC therapy was MTB-recommended to all 39 patients with actionable/unexploited alterations, but only 22 (56%) accessed the applicable drug, mainly due to clinical deterioration, lengthy drug-gathering procedures, and geographical distance from recruiting clinical trials. Partial response and stable disease were recorded in 8 and 7 of 19 evaluable patients, respectively. The time to progression (TTP) ratio (MTB-recommended treatment vs last pre-MTB treatment) exceeded the conventional Von Hoff 1.3 cut-off in 9/19 cases, high absolute TTP and Von Hoff values coinciding in 3 cases. Retrospectively, 8 patients receiving post-MTB treatment(s) as per physician's choice were noted to have a much longer overall survival from MTB accrual than 11 patients who had received no further treatment (35.09 vs 6.67 months, p = 0.006). CONCLUSIONS: MTB-recommended/non-SoC treatments are effective, including those assigned by ctDNA-only alterations. However, real-world MTBs may inadvertently recruit patients electively susceptible to diverse and/or multiple treatments.

Development of a somatic variant registry in a National Cancer Center: towards Molecular Real World Data preparedness.

The Biomedical Research field is currently advancing to develop Clinical Trials and translational projects based on Real World Evidence. To make this transition feasible, clinical centers need to work toward Data Accessibility and Interoperability. This task is particularly challenging when applied to Genomics, that entered in routinary screening in the last years via mostly amplicon-based Next-Generation Sequencing panels. Said experiments produce up to hundreds of features per patient, and their summarized results are often stored in static clinical reports, making critical information inaccessible to automated access and Federated Search consortia. In this study, we present a reanalysis of 4620 solid tumor sequencing samples in five different histology settings. Furthermore, we describe all the Bioinformatics and Data Engineering processes that were put in place in order to create a Somatic Variant Registry able to deal with the large biotechnological variability of routinary Genomics Profiling.

Machine learning endometrial cancer risk prediction model: integrating guidelines of European Society for Medical Oncology with the tumor immune framework.

OBJECTIVE: Current prognostic factors for endometrial cancer are not sufficient to predict recurrence in early stages. Treatment choices are based on the prognostic factors included in the risk classes defined by the ESMO-ESGO-ESTRO (European Society for Medical Oncology-European Society of Gynaecological Oncology-European Society for Radiotherapy and Oncology) consensus conference with the new biomolecular classification based on POLE, TP53, and microsatellite instability status. However, a minority of early stage cases relapse regardless of their low risk profiles. Integration of the immune context status to existing molecular based models has not been fully evaluated. This study aims to investigate whether the integration of the immune landscape in the tumor microenvironment could improve clinical risk prediction models and allow better profiling of early stages. METHODS: Leveraging the potential of in silico deconvolution tools, we estimated the relative abundances of immune populations in public data and then applied feature selection methods to generate a machine learning based model for disease free survival probability prediction. RESULTS: We included information on International Federation of Gynecology and Obstetrics (FIGO) stage, tumor mutational burden, microsatellite instability, POLEmut status, interferon γ signature, and relative abundances of monocytes, natural killer cells, and CD4+T cells to build a relapse prediction model and obtained a balanced accuracy of 69%. We further identified two novel early stage profiles that undergo different pathways of recurrence. CONCLUSION: This study presents an extension of current prognostic factors for endometrial cancer by exploiting machine learning models and deconvolution techniques on available public biomolecular data. Prospective clinical trials are advisable to validate the early stage stratification.

Biomolecular Classification in Endometrial Cancer: Onset, Evolution, and Further Perspectives: A Critical Review.

Since the new guidelines for endometrial cancer risk classification have been published, many reviews have proposed a critical re-evaluation. In this review, we look back to how the molecular classification system was built and its evolution in time to highlight the major flaws, particularly the biases stemming from the inherent limitations of the cohorts involved in the discoveries. A significant drawback in some cohorts is the inclusion criteria, as well as the retrospective nature and the notably sparse numbers, especially in the POLEmut (nonsynonymous mutation in EDM domain of POLE) risk groups, all of which impact the reliability of outcomes. Additionally, a disregard for variations in follow-up duration leads to a non-negligible bias, which raises a substantial concern in data interpretation and guideline applicability. Finally, according to the results that we obtained through a re-analysis of the confirmation cohort, the p53abn (IHC positive for p53 protein) subgroup, which is predominant in non-endometrioid histology (73-80%), loses its predictivity power in the endometrioid cohort of patients. The exclusion of non-endometrioid subtypes from the cohort led to a complete overlap of three molecular subgroups (all except POLEmut) for both overall and progression-free survival outcomes, suggesting the need for a more histotype-specific approach. In conclusion, this review challenges the current ESGO/ESTRO/ESP guidelines on endometrial cancer risk classification and highlights the limitations that must be addressed to better guide the clinical decision-making process.

AI drives the assessment of lung cancer microenvironment composition.

Purpose: The abundance and distribution of tumor-infiltrating lymphocytes (TILs) as well as that of other components of the tumor microenvironment is of particular importance for predicting response to immunotherapy in lung cancer (LC). We describe here a pilot study employing artificial intelligence (AI) in the assessment of TILs and other cell populations, intending to reduce the inter- or intra-observer variability that commonly characterizes this evaluation. Design: We developed a machine learning-based classifier to detect tumor, immune, and stromal cells on hematoxylin and eosin-stained sections, using the open-source framework QuPath. We evaluated the quantity of the aforementioned three cell populations among 37 LC whole slide images regions of interest, comparing the assessments made by five pathologists, both before and after using graphical predictions made by AI, for a total of 1110 quantitative measurements. Results: Our findings indicate noteworthy variations in score distribution among pathologists and between individual pathologists and AI. The AI-guided pathologist's evaluations resulted in reduction of significant discrepancies across pathologists: three comparisons showed a loss of significance (p > 0.05), whereas other four showed a reduction in significance (p > 0.01). Conclusions: We show that employing a machine learning approach in cell population quantification reduces inter- and intra-observer variability, improving reproducibility and facilitating its use in further validation studies.

Towards reproducible research in recurrent pregnancy loss immunology: Learning from cancer microenvironment deconvolution.

The most recent international guidelines regarding recurrent pregnancy loss (RPL) exclude most of the immunological tests recommended for RPL since they do not reach an evidence-based level. Comparisons for metanalysis and systematic reviews are limited by the ambiguity in terms of RPL definition, etiological and risk factors, diagnostic work-up, and treatments applied. Therefore, cohort heterogeneity, the inadequacy of numerosity, and the quality of data confirm a not standardized research quality in the RPL field, especially for immunological background, for which potential research application remains confined in a separate single biological layer. Innovative sequencing technologies and databases have proved to play a significant role in the exploration and validation of cancer research in the context of dataset quality and bioinformatics tools. In this article, we will investigate how bioinformatics tools born for large-scale cancer immunological research could revolutionize RPL immunological research but are limited by the nature of current RPL datasets.

Optimizing the Illumina COVIDSeq laboratorial and bioinformatics pipeline on thousands of samples for SARS-CoV-2 Variants of Concern tracking.

The SARS-CoV-2 Variants of Concern tracking via Whole Genome Sequencing represents a pillar of public health measures for the containment of the pandemic. The ability to track down the lineage distribution on a local and global scale leads to a better understanding of immune escape and to adopting interventions to contain novel outbreaks. This scenario poses a challenge for NGS laboratories worldwide that are pressed to have both a faster turnaround time and a high-throughput processing of swabs for sequencing and analysis. In this study, we present an optimization of the Illumina COVID-seq protocol carried out on thousands of SARS-CoV-2 samples at the wet and dry level. We discuss the unique challenges related to processing hundreds of swabs per week such as the tradeoff between ultra-high sensitivity and negative contamination levels, cost efficiency and bioinformatics quality metrics.