The survival benefit with adjuvant chemotherapy for patients with resected stage II-III non-small-cell lung cancer (NSCLC) is modest. Efforts to develop prognostic or predictive biomarkers in these ...patients have not yielded clinically useful tests. We report findings from the Lung Adjuvant Cisplatin Evaluation (LACE)-Bio-II study, in which we analyzed next-generation sequencing and long-term outcomes data from > 900 patients with early-stage NSCLC treated prospectively in adjuvant landmark clinical trials. We used a targeted gene panel to assess the prognostic and predictive effect of mutations in individual genes, DNA repair pathways, and tumor mutation burden (TMB).
A total of 908 unmatched, formalin-fixed, paraffin-embedded, resected lung cancer tumor specimens were sequenced using a targeted panel of 1,538 genes. Stringent filtering criteria were applied to exclude germline variants and artifacts related to formalin fixation. Disease-free survival, overall survival, and lung cancer-specific survival (LCSS) were assessed in Cox models stratified by trial and adjusted for treatment, age, sex, performance score, histology, type of surgery, and stage.
Nonsynonymous mutations were identified in 1,515 genes in 908 tumor samples. High nonsynonymous TMB (> 8 mutations/Mb) was prognostic for favorable outcomes (ie, overall survival, disease-free survival, and LCSS) in patients with resected NSCLC. LCSS benefit with adjuvant chemotherapy was more pronounced in patients with low nonsynonymous TMBs (≤ 4 mutations/Mb). Presence of mutations in DNA repair pathways, tumor-infiltrating lymphocytes, TP53 alteration subtype, and intratumor heterogeneity was neither prognostic nor predictive. Statistically significant effect of mutations in individual genes was difficult to determine due to high false-discovery rates.
High nonsynonymous TMB was associated with a better prognosis in patients with resected NSCLC. In addition, the benefit of adjuvant chemotherapy on LCSS was more pronounced in patients with low nonsynonymous TMBs. Studies are warranted to confirm these findings.
We undertook this analysis of KRAS mutation in four trials of adjuvant chemotherapy (ACT) versus observation (OBS) to clarify the prognostic/predictive roles of KRAS in non-small-cell lung cancer ...(NSCLC).
KRAS mutation was determined in blinded fashion. Exploratory analyses were performed to characterize relationships between mutation status and subtype and survival outcomes using a multivariable Cox model.
Among 1,543 patients (763 OBS, 780 ACT), 300 had KRAS mutations (codon 12, n = 275; codon 13, n = 24; codon 14, n = 1). In OBS patients, there was no prognostic difference for overall survival for codon-12 (mutation v wild type WT hazard ratio HR = 1.04; 95% CI, 0.77 to 1.40) or codon-13 (HR = 1.01; 95% CI, 0.47 to 2.17) mutations. No significant benefit from ACT was observed for WT-KRAS (ACT v OBS HR = 0.89; 95% CI, 0.76 to 1.04; P = .15) or codon-12 mutations (HR = 0.95; 95% CI, 0.67 to 1.35; P = .77); with codon-13 mutations, ACT was deleterious (HR = 5.78; 95% CI, 2.06 to 16.2; P < .001; interaction P = .002). There was no prognostic effect for specific codon-12 amino acid substitution. The effect of ACT was variable among patients with codon-12 mutations: G12A or G12R (HR = 0.66; P = .48), G12C or G12V (HR = 0.94; P = .77) and G12D or G12S (HR = 1.39; P = .48; comparison of four HRs, including WT, interaction P = .76). OBS patients with KRAS-mutated tumors were more likely to develop second primary cancers (HR = 2.76, 95% CI, 1.34 to 5.70; P = .005) but not ACT patients (HR = 0.66; 95% CI, 0.25 to 1.75; P = .40; interaction, P = .02).
KRAS mutation status is not significantly prognostic. The potential interaction in patients with codon-13 mutations requires validation. At this time, KRAS status cannot be recommended to select patients with NSCLC for ACT.
Network meta‐analysis (NMA) allows the combination of direct and indirect evidence from a set of randomized clinical trials. Performing NMA using individual patient data (IPD) is considered as a ...“gold standard” approach as it provides several advantages over NMA based on aggregate data. For example, it allows to perform advanced modeling of covariates or covariate‐treatment interactions. An important issue in IPD NMA is the selection of influential parameters among terms that account for inconsistency, covariates, covariate‐by‐treatment interactions or nonproportionality of treatments effect for time to event data. This issue has not been deeply studied in the literature yet and in particular not for time‐to‐event data. A major difficulty is to jointly account for between‐trial heterogeneity which could have a major influence on the selection process. The use of penalized generalized mixed effect model is a solution, but existing implementations have several shortcomings and an important computational cost that precludes their use for complex IPD NMA. In this article, we propose a penalized Poisson regression model to perform IPD NMA of time‐to‐event data. It is based only on fixed effect parameters which improve its computational cost over the use of random effects. It could be easily implemented using existing penalized regression package. Computer code is shared for implementation. The methods were applied on simulated data to illustrate the importance to take into account between trial heterogeneity during the selection procedure. Finally, it was applied to an IPD NMA of overall survival of chemotherapy and radiotherapy in nasopharyngeal carcinoma.
Tumor lymphocytic infiltration (TLI) has differing prognostic value among various cancers. The objective of this study was to assess the effect of TLI in lung cancer.
A discovery set (one trial, n = ...824) and a validation set (three trials, n = 984) that evaluated the benefit of platinum-based adjuvant chemotherapy in non-small-cell lung cancer were used as part of the LACE-Bio (Lung Adjuvant Cisplatin Evaluation Biomarker) study. TLI was defined as intense versus nonintense. The main end point was overall survival (OS); secondary end points were disease-free survival (DFS) and specific DFS (SDFS). Hazard ratios (HRs) and 95% CIs associated with TLI were estimated through a multivariable Cox model in both sets. TLI-histology and TLI-treatment interactions were explored in the combined set.
Discovery and validation sets with complete data included 783 (409 deaths) and 763 (344 deaths) patients, respectively. Median follow-up was 4.8 and 6 years, respectively. TLI was intense in 11% of patients in the discovery set compared with 6% in the validation set (P < .001). The prognostic value of TLI in the discovery set (OS: HR, 0.56; 95% CI, 0.38 to 0.81; P = .002; DFS: HR, 0.59; 95% CI, 0.42 to 0.83; P = .002; SDFS: HR, 0.56; 95% CI, 0.38 to 0.82; P = .003) was confirmed in the validation set (OS: HR, 0.45; 95% CI, 0.23 to 0.85; P = .01; DFS: HR, 0.44; 95% CI, 0.24 to 0.78; P = .005; SDFS: HR, 0.42; 95% CI, 0.22 to 0.80; P = .008) with no heterogeneity across trials (P ≥ .38 for all end points). No significant predictive effect was observed for TLI (P ≥ .78 for all end points).
Intense lymphocytic infiltration, found in a minority of tumors, was validated as a favorable prognostic marker for survival in resected non-small-cell lung cancer.
The research of biomarker-treatment interactions is commonly investigated in randomized clinical trials (RCT) for improving medicine precision. The hierarchical interaction constraint states that an ...interaction should only be in a model if its main effects are also in the model. However, this constraint is not guaranteed in the standard penalized statistical approaches. We aimed to find a compromise for high-dimensional data between the need for sparse model selection and the need for the hierarchical constraint.
To favor the property of the hierarchical interaction constraint, we proposed to create groups composed of the biomarker main effect and its interaction with treatment and to perform the bi-level selection on these groups. We proposed two weighting approaches (Single Wald (SW) and likelihood ratio test (LRT)) for the adaptive lasso method. The selection performance of these two approaches is compared to alternative lasso extensions (adaptive lasso with ridge-based weights, composite Minimax Concave Penalty, group exponential lasso and Sparse Group Lasso) through a simulation study. A RCT (NSABP B-31) randomizing 1574 patients (431 events) with early breast cancer aiming to evaluate the effect of adjuvant trastuzumab on distant-recurrence free survival with expression data from 462 genes measured in the tumour will serve for illustration. The simulation study illustrates that the adaptive lasso LRT and SW, and the group exponential lasso favored the hierarchical interaction constraint. Overall, in the alternative scenarios, they had the best balance of false discovery and false negative rates for the main effects of the selected interactions. For NSABP B-31, 12 gene-treatment interactions were identified more than 20% by the different methods. Among them, the adaptive lasso (SW) approach offered the best trade-off between a high number of selected gene-treatment interactions and a high proportion of selection of both the gene-treatment interaction and its main effect.
Adaptive lasso with Single Wald and likelihood ratio test weighting and the group exponential lasso approaches outperformed their competitors in favoring the hierarchical constraint of the biomarker-treatment interaction. However, the performance of the methods tends to decrease in the presence of prognostic biomarkers.
The classification for invasive lung adenocarcinoma by the International Association for the Study of Lung Cancer, American Thoracic Society, European Respiratory Society, and WHO is based on the ...predominant histologic pattern-lepidic (LEP), papillary (PAP), acinar (ACN), micropapillary (MIP), or solid (SOL)-present in the tumor. This classification has not been tested in multi-institutional cohorts or clinical trials or tested for its predictive value regarding survival from adjuvant chemotherapy (ACT).
Of 1,766 patients in the IALT, JBR.10, CALGB 9633 (Alliance), and ANITA ACT trials included in the LACE-Bio study, 725 had adenocarcinoma. Histologies were reclassified according to the new classification and then collapsed into three groups (LEP, ACN/PAP, and MIP/SOL). Primary end point was overall survival (OS); secondary end points were disease-free survival (DFS) and specific DFS (SDFS). Hazard ratios (HRs) and 95% CIs were estimated through multivariable Cox models stratified by trial. Prognostic value was estimated in the observation arm and predictive value by a treatment effect interaction with histologic subgroups. Significance level was set at .01 for pooled analysis.
A total of 575 patients were included in this analysis. OS was not prognostically different between histologic subgroups, but univariable DFS and SDFS were worse for MIP/SOL compared with LEP or ACN/PAP subgroup (P < .01); this remained marginally significant after adjustment. MIP/SOL patients (but not ACN/PAP) derived DFS and SDFS but not OS benefit from ACT (OS: HR, 0.71; 95% CI, 0.51 to 0.99; interaction P = .18; DFS: HR, 0.60; 95% CI, 0.44 to 0.82; interaction P = < .01; and SDFS: HR, 0.59; 95% CI, 0.42 to 0.81; interaction P = .01).
The new lung adenocarcinoma classification based on predominant histologic pattern was not predictive for ACT benefit for OS, but it seems predictive for disease-specific outcomes.
Background Diagnosis of COVID-19 in symptomatic patients and screening of populations for SARS-CoV-2 infection require access to straightforward, low-cost and high-throughput testing. The recommended ...nasopharyngeal swab tests are limited by the need of trained professionals and specific consumables and this procedure is poorly accepted as a screening method In contrast, saliva sampling can be self-administered. Methods In order to compare saliva and nasopharyngeal/oropharyngeal samples for the detection of SARS-CoV-2, we designed a meta-analysis searching in PubMed up to December 29th, 2020 with the key words "(SARS-CoV-2 OR COVID-19 OR COVID19) AND (salivary OR saliva OR oral fluid)) NOT (reviewPublication Type) NOT (PrePrintPublication Type)" applying the following criteria: records published in peer reviewed scientific journals, in English, with at least 15 nasopharyngeal/orapharyngeal swabs and saliva paired samples tested by RT-PCR, studies with available raw data including numbers of positive and negative tests with the two sampling methods. For all studies, concordance and sensitivity were calculated and then pooled in a random-effects model. Findings A total of 377 studies were retrieved, of which 50 were eligible, reporting on 16,473 pairs of nasopharyngeal/oropharyngeal and saliva samples. Meta-analysis showed high concordance, 92.5% (95%CI: 89.5-94.7), across studies and pooled sensitivities of 86.5% (95%CI: 83.4-89.1) and 92.0% (95%CI: 89.1-94.2) from saliva and nasopharyngeal/oropharyngeal swabs respectively. Heterogeneity across studies was 72.0% for saliva and 85.0% for nasopharyngeal/oropharyngeal swabs. Interpretation Our meta-analysis strongly suggests that saliva could be used for frequent testing of COVID-19 patients and "en masse" screening of populations.
Purpose Our previous work evaluated individual prognostic and predictive roles of TP53, KRAS, and EGFR in non-small-cell lung cancer (NSCLC). In this analysis, we explore the prognostic and ...predictive roles of TP53/KRAS and TP53/EGFR comutations in randomized trials of adjuvant chemotherapy versus observation. Patients and Methods Mutation analyses (wild-type WT and mutant) for TP53, KRAS, and EGFR were determined in blinded fashion in multiple laboratories. Primary and secondary end points of pooled analysis were overall survival and disease-free survival. We evaluated the role of TP53/KRAS comutation in all patients and in the adenocarcinoma subgroup as well as the TP53/EGFR comutation in adenocarcinoma only through a multivariable Cox proportional hazards model stratified by trial. Results Of 3,533 patients with NSCLC, 1,181 (557 deaths) and 404 (170 deaths) were used for TP53/KRAS and TP53/EGFR analyses. For TP53/KRAS mutation status, no prognostic effect was observed ( P = .61), whereas a borderline predictive effect ( P = .04) was observed with a deleterious effect of chemotherapy with TP53/KRAS comutations versus WT/WT (hazard ratio, 2.49 95% CI, 1.10 to 5.64; P = .03). TP53/EGFR comutation in adenocarcinoma was neither prognostic ( P = .83), nor significantly predictive ( P = .86). Similar results were observed for both groups for disease-free survival. Conclusion We could identify no prognostic effect of the KRAS or EGFR driver and TP53 tumor suppressor comutation. Our observation of a potential negative predictive effect of TP53/KRAS comutation requires validation.
Individual patient data (IPD) present particular advantages in network meta-analysis (NMA) because interactions may lead an aggregated data (AD)-based model to wrong a treatment effect (TE) ...estimation. However, fewer works have been conducted for IPD with time-to-event contrary to binary outcomes. We aimed to develop a general frequentist one-step model for evaluating TE in the presence of interaction in a three-node NMA for time-to-event data.
One-step, frequentist, IPD-based Cox and Poisson generalized linear mixed models were proposed. We simulated a three-node network with or without a closed loop with (1) no interaction, (2) covariate-treatment interaction, and (3) covariate distribution heterogeneity and covariate-treatment interaction. These models were applied to the NMA (Meta-analyses of Chemotherapy in Head and Neck Cancer MACH-NC and Radiotherapy in Carcinomas of Head and Neck MARCH), which compared the addition of chemotherapy or modified radiotherapy (mRT) to loco-regional treatment with two direct comparisons. AD-based (contrast and meta-regression) models were used as reference.
In the simulated study, no IPD models failed to converge. IPD-based models performed well in all scenarios and configurations with small bias. There were few variations across different scenarios. In contrast, AD-based models performed well when there were no interactions, but demonstrated some bias when interaction existed and a larger one when the modifier was not distributed evenly. While meta-regression performed better than contrast-based only, it demonstrated a large variability in estimated TE. In the real data example, Cox and Poisson IPD-based models gave similar estimations of the model parameters. Interaction decomposition permitted by IPD explained the ecological bias observed in the meta-regression.
The proposed general one-step frequentist Cox and Poisson models had small bias in the evaluation of a three-node network with interactions. They performed as well or better than AD-based models and should also be undertaken whenever possible.