Nils Ternès

Assessing tumor-infiltrating lymphocytes in solid tumors: a practical review for pathologists and proposal for a standardized method from the International Immuno-Oncology Biomarkers Working Group: part 2: TILs in melanoma, gastrointestinal tract carcinomas, non-small cell lung carcinoma and mesothelioma, endometrial and ovarian carcinomas, squamous cell carcinoma of the head and neck, genitourinary carcinomas, and primary brain Tumors

Assessment of the immune response to tumors is growing in importance as the prognostic implications of this response are increasingly recognized, and as immunotherapies are evaluated and implemented in different tumor types. However, many different approaches can be used to assess and describe the immune response, which limits efforts at implementation as a routine clinical biomarker. In part 1 of this review, we have proposed a standardized methodology to assess tumor-infiltrating lymphocytes (TILs) in solid tumors, based on the International Immuno-Oncology Biomarkers Working Group guidelines for invasive breast carcinoma. In part 2 of this review, we discuss the available evidence for the prognostic and predictive value of TILs in common solid tumors, including carcinomas of the lung, gastrointestinal tract, genitourinary system, gynecologic system, and head and neck, as well as primary brain tumors, mesothelioma and melanoma. The particularities and different emphases in TIL assessment in different tumor types are discussed. The standardized methodology we propose can be adapted to different tumor types and may be used as a standard against which other approaches can be compared. Standardization of TIL assessment will help clinicians, researchers and pathologists to conclusively evaluate the utility of this simple biomarker in the current era of immunotherapy.

Update on tumor-infiltrating lymphocytes (TILs) in breast cancer, including recommendations to assess TILs in residual disease after neoadjuvant therapy and in carcinoma in situ: A report of the International Immuno-Oncology Biomarker Working Group on Breast Cancer

Morphological evaluation of tumor-infiltrating lymphocytes (TILs) in breast cancer is gaining momentum as evidence strengthens the clinical relevance of this immunological biomarker. TILs in the post-neoadjuvant residual disease setting are acquiring increasing importance as a stratifying marker in clinical trials, considering the raising interest on immunotherapeutic strategies after neoadjuvant chemotherapy. TILs in ductal carcinoma in situ, with or without invasive carcinoma, represent an emerging area of clinical breast cancer research. The aim of this report is to update pathologists, clinicians and researchers on TIL assessment in both the post-neoadjuvant residual disease and the ductal carcinoma in situ settings. The International Immuno-Oncology Working Group proposes a method for assessing TILs in these settings, based on the previously published International Guidelines on TIL Assessment in Breast Cancer. In this regard, these recommendations represent a consensus guidance for pathologists, aimed to achieve the highest possible consistency among future studies.

Empirical extensions of the lasso penalty to reduce the false discovery rate in high-dimensional Cox regression models.

Correct selection of prognostic biomarkers among multiple candidates is becoming increasingly challenging as the dimensionality of biological data becomes higher. Therefore, minimizing the false discovery rate (FDR) is of primary importance, while a low false negative rate (FNR) is a complementary measure. The lasso is a popular selection method in Cox regression, but its results depend heavily on the penalty parameter λ. Usually, λ is chosen using maximum cross-validated log-likelihood (max-cvl). However, this method has often a very high FDR. We review methods for a more conservative choice of λ. We propose an empirical extension of the cvl by adding a penalization term, which trades off between the goodness-of-fit and the parsimony of the model, leading to the selection of fewer biomarkers and, as we show, to the reduction of the FDR without large increase in FNR. We conducted a simulation study considering null and moderately sparse alternative scenarios and compared our approach with the standard lasso and 10 other competitors: Akaike information criterion (AIC), corrected AIC, Bayesian information criterion (BIC), extended BIC, Hannan and Quinn information criterion (HQIC), risk information criterion (RIC), one-standard-error rule, adaptive lasso, stability selection, and percentile lasso. Our extension achieved the best compromise across all the scenarios between a reduction of the FDR and a limited raise of the FNR, followed by the AIC, the RIC, and the adaptive lasso, which performed well in some settings. We illustrate the methods using gene expression data of 523 breast cancer patients. In conclusion, we propose to apply our extension to the lasso whenever a stringent FDR with a limited FNR is targeted. Copyright

Statistical controversies in clinical research: prognostic gene signatures are not (yet) useful in clinical practice

With the genomic revolution and the era of targeted therapy, prognostic and predictive gene signatures are becoming increasingly important in clinical research. They are expected to assist prognosis assessment and therapeutic decision making. Notwithstanding, an evidence-based approach is needed to bring gene signatures from the laboratory to clinical practice. In early breast cancer, multiple prognostic gene signatures are commercially available without having formally reached the highest levels of evidence-based criteria. We discuss specific concepts for developing and validating a prognostic signature and illustrate them with contemporary examples in breast cancer. When a prognostic signature has not been developed for predicting the magnitude of relative treatment benefit through an interaction effect, it may be wishful thinking to test its predictive value. We propose that new gene signatures be built specifically for predicting treatment effects for future patients and outline an approach for this using a cross-validation scheme in a standard phase III trial. Replication in an independent trial remains essential.

Identification of biomarker‐by‐treatment interactions in randomized clinical trials with survival outcomes and high‐dimensional spaces

Stratified medicine seeks to identify biomarkers or parsimonious gene signatures distinguishing patients that will benefit most from a targeted treatment. We evaluated 12 approaches in high‐dimensional Cox models in randomized clinical trials: penalization of the biomarker main effects and biomarker‐by‐treatment interactions (full‐lasso, three kinds of adaptive lasso, ridge+lasso and group‐lasso); dimensionality reduction of the main effect matrix via linear combinations (PCA+lasso (where PCA is principal components analysis) or PLS+lasso (where PLS is partial least squares)); penalization of modified covariates or of the arm‐specific biomarker effects (two‐I model); gradient boosting; and univariate approach with control of multiple testing. We compared these methods via simulations, evaluating their selection abilities in null and alternative scenarios. We varied the number of biomarkers, of nonnull main effects and true biomarker‐by‐treatment interactions. We also proposed a novel measure evaluating the interaction strength of the developed gene signatures. In the null scenarios, the group‐lasso, two‐I model, and gradient boosting performed poorly in the presence of nonnull main effects, and performed well in alternative scenarios with also high interaction strength. The adaptive lasso with grouped weights was too conservative. The modified covariates, PCA+lasso, PLS+lasso, and ridge+lasso performed moderately. The full‐lasso and adaptive lassos performed well, with the exception of the full‐lasso in the presence of only nonnull main effects. The univariate approach performed poorly in alternative scenarios. We also illustrate the methods using gene expression data from 614 breast cancer patients treated with adjuvant chemotherapy.

Statistical methods applied to omics data: predicting response to neoadjuvant therapy in breast cancer.

Purpose of review Omics technologies have become an essential part of clinical trials in oncology to provide a better understanding of molecular mechanisms and to unveil therapeutic targets. Standard statistical methods often fail in the high-dimensional setting. Therefore, an adequate modelling of the omics data is needed in order to identify ‘target’ genes of interest. Recent findings Several genes or gene signatures have been identified to predict the response to neoadjuvant therapies in breast cancer trials. We first reviewed statistical methods used to identify genes in 13 recent publications. Most of these studies had a small sample size (median: 42 patients) and were nonrandomized. We then focused on some popular methods – especially the so-called penalized methods used by three of the reviewed articles – and on the more recent methods proposed to predict causal estimates from observational data. We finally illustrated these methods in a nonrandomized neoadjuvant phase II trial of letrozole in estrogen receptor-positive breast cancer patients. Summary The review highlighted small sample sizes, few randomized trials and a large panel of statistical methods used in this setting. In our illustrated neoadjuvant example, causal inference methods did not outperform the penalized methods.

biospear: an R package for biomarker selection in penalized Cox regression

Summary The R package biospear allows selecting the biomarkers with the strongest impact on survival and on the treatment effect in high-dimensional Cox models, and estimating expected survival probabilities. Most of the implemented approaches are based on penalized regression techniques. Availability and implementation The package is available on the CRAN. (https://CRAN.R-project.org/package=biospear). Contact stefan.michiels@gustaveroussy.fr. Supplementary information Supplementary data are available at Bioinformatics online.

Contour plot assessment of existing meta-analyses confirms robust association of statin use and acute kidney injury risk

OBJECTIVES Robustness of an existing meta-analysis can justify decisions on whether to conduct an additional study addressing the same research question. We illustrate the graphical assessment of the potential impact of an additional study on an existing meta-analysis using published data on statin use and the risk of acute kidney injury. STUDY DESIGN AND SETTING A previously proposed graphical augmentation approach is used to assess the sensitivity of the current test and heterogeneity statistics extracted from existing meta-analysis data. In addition, we extended the graphical augmentation approach to assess potential changes in the pooled effect estimate after updating a current meta-analysis and applied the three graphical contour definitions to data from meta-analyses on statin use and acute kidney injury risk. RESULTS In the considered example data, the pooled effect estimates and heterogeneity indices demonstrated to be considerably robust to the addition of a future study. Supportingly, for some previously inconclusive meta-analyses, a study update might yield statistically significant kidney injury risk increase associated with higher statin exposure. CONCLUSIONS The illustrated contour approach should become a standard tool for the assessment of the robustness of meta-analyses. It can guide decisions on whether to conduct additional studies addressing a relevant research question.

Prediction of treatment benefit in high-dimensional cox models via gene signatures in randomized clinical trials

Methods We investigated four approaches: penalize biomarker main effects and biomarker-by-treatment interactions using a lasso penalty (full-lasso); control of main effects by principal components or ridge penalty, and lasso on interactions (sPCA+lasso or ridge+lasso); and ‘modified covariates’ in a penalized regression model (Tian et al. 2014). We performed simulations under null and alternative scenarios by varying the sample size n, number of biomarkers H, number of true main effects or treatment-modifiers, effect sizes and correlations. We proposed two novel measures of treatment effect prediction for gene signatures: a difference in C-indices and a Wald-based interaction statistic. We used gene expression data from a RCT of adjuvant chemotherapy in non-small cell lung cancer (n=133) for illustration.