Citations

612 Background: mCRC is a heterogeneous disease leading to possible disparate responses among lesions in a single pt. This study assesses the heterogeneity of lesion responses and evaluates the prognostic value of LBR12 and its mortality risk discrimination beyond RECIST responses. Methods: Pts with ≥ 2 lesions and no new lesions at 12 weeks (wks) were eligible. For each pt, after 12 wks of Rx, each lesion was categorized as progressive disease (L-PD: ≥20% increase in the longest diameter [LD]), partial response (L-PR: ≥30% decrease in LD), or stable disease (L-SD: neither L-PD nor L-PR). LBR12 was defined, per patient, by the combination of lesion responses: homogeneous LBR12 (L-PD, L-SD, L-PR only) and heterogeneous LBR12 (mixture of L-PD/L-SD, L-PD/L-PR, L-SD/L-PR). LBR12 and overall survival (OS) were correlated using stratified multivariate Cox models after adjusting for age, gender, and ECOG PS. Results: Among 9,092 mCRCs (Rx: chemo alone 44%; chemo + VEGF inhibitor [VEGFi] 42%; chemo + EGFR inhibitor [EGFRi] 10%) from 16 1st-line studies. Median OS: 2.2 years. Per RECIST at 12 wks, CR 0%; PR 36.1%; SD 60.9%; PD 3.0%. Responses in 52% were heterogeneous (Table). VEGFi and EGFRi treated pts had the highest rate of L-SD/L-PR (45%) and L-PR only (22%) status, respectively. Median OS increased monotonically across pts with more L-PRs and fewer L-PDs (Table). Pts with L-SD/L-PR status, among which 51% had SD per RECIST, had longer OS than those with L-SD only status (HRadj.= .81, padj.< .0001), but shorter OS than those with L-PR only status (HRadj.= 1.46, padj.< .0001). Pts with L-PD/L-SD status, among which 71% were SD per RECIST, had shorter OS than those with L-SD only status (HRadj.= 2.22, padj.< .0001). These associations were consistent across treatment regimens. Conclusions: The lesion-based response captures the heterogeneity of within pt lesion responses and provides refinement in predicting outcome beyond RECIST response at 12 wks. [Table: see text]

3535 Background: CRC is known to be a heterogeneous disease. This study quantifies within pt heterogeneity in early lesion change rate (LCR) in the era of targeted agents compared to chemo alone and its potential impact on survival. Methods: Pts with 2-10 lesions measured at baseline were included. For each lesion, the early LCR was defined as the change in size (shrinking or growing) from baseline to 12 weeks on treatment. Within pt heterogeneity in early LCR among lesions was estimated by standard deviation (STD). A larger value of STD indicates larger variation of LCR per pt. Stratified multivariate Cox models were used to assess the associations between LCR STD with overall survival (OS). Adjusted hazard ratios (HRadj) and 95% confidence intervals (CIs) are reported. Results: Data were available on 9,092 mCRC pts (median age 61; 60% male, 55% ECOG PS 0; 61% 2+ metastatic sites) enrolled in 16 1st-line randomized trials, with 44%, 42%, and 10% of pts received chemo alone, + a VEGF inhibitor (VEGFi) or an EGFR inhibitor (EGFRi), respectively. LCR heterogeneity is the highest among pts received EGFRi but lowest among pts received VEGFi (Table). Overall, higher heterogeneity is associated with worst OS (HRadj1.22, 95% CI (1.16, 1.27)). The effect is most pronounced in pts received VEGFi (interaction p=0.0012). Conclusions: There was heterogeneity observed in lesion size changes within pts. Its magnitude varies across treatment approaches, and was associated with poor survival. This preliminary result reveals the great potentials to define novel response endpoint and refine treatment decision-making by incorporating heterogeneities in lesion changes. [Table: see text]

Abstract “I have noticed in operations of this kind, which I have seen performed by others upon the living, and in a number of excisions, which I have myself performed on the dead body, that most of the difficulty in the separation of the tumor has occurred in the region of these ligaments…. This difficulty, I believe, to be a very frequent source of that accident, which so commonly occurs in removal of goiter, I mean division of the recurrent laryngeal nerve.” Sir James Berry (1887)

Abstract Motivation Recent studies have revealed a complex interplay between environment, the human microbiome and health and disease. Mediation analysis of the human microbiome in these complex relationships could potentially provide insights into the role of the microbiome in the etiology of disease and, more importantly, lead to novel clinical interventions by modulating the microbiome. However, due to the high dimensionality, sparsity, non-normality and phylogenetic structure of microbiome data, none of the existing methods are suitable for testing such clinically important mediation effect. Results We propose a distance-based approach for testing the mediation effect of the human microbiome. In the framework, the nonlinear relationship between the human microbiome and independent/dependent variables is captured implicitly through the use of sample-wise ecological distances, and the phylogenetic tree information is conveniently incorporated by using phylogeny-based distance metrics. Multiple distance metrics are utilized to maximize the power to detect various types of mediation effect. Simulation studies demonstrate that our method has correct Type I error control, and is robust and powerful under various mediation models. Application to a real gut microbiome dataset revealed that the association between the dietary fiber intake and body mass index was mediated by the gut microbiome. Availability and implementation An R package ‘MedTest’ is freely available at https://github.com/jchen1981/MedTest. Supplementary information Supplementary data are available at Bioinformatics online.

AbstractMotivationOne objective of human microbiome studies is to identify differentially abundant microbes across biological conditions. Previous statistical methods focus on detecting the shift in the abundance and/or prevalence of the microbes and treat the dispersion (spread of the data) as a nuisance. These methods also assume that the dispersion is the same across conditions, an assumption which may not hold in presence of sample heterogeneity. Moreover, the widespread outliers in the microbiome sequencing data make existing parametric models not overly robust. Therefore, a robust and powerful method that allows covariate-dependent dispersion and addresses outliers is still needed for differential abundance analysis.ResultsWe introduce a novel test for differential distribution analysis of microbiome sequencing data by jointly testing the abundance, prevalence and dispersion. The test is built on a zero-inflated negative binomial regression model and winsorized count data to account for zero-inflation and outliers. Using simulated data and real microbiome sequencing datasets, we show that our test is robust across various biological conditions and overall more powerful than previous methods.Availability and implementationR package is available at https://github.com/jchen1981/MicrobiomeDDA.Supplementary informationSupplementary data are available at Bioinformatics online.