Efficient Identification of Patients With NTRK Fusions Using a Supervised Tumor-Agnostic Approach.

CONTEXT.­: The neurotrophic tropomyosin receptor kinase (NTRK) family gene rearrangements have been recently incorporated as predictive biomarkers in a "tumor-agnostic" manner. However, the identification of these patients is extremely challenging because the overall frequency of NTRK fusions is below 1%. Academic groups and professional organizations have released recommendations on the algorithms to detect NTRK fusions. The European Society for Medical Oncology proposal encourages the use of next-generation sequencing (NGS) if available, or alternatively immunohistochemistry (IHC) could be used for screening with NGS confirmation of all positive IHC results. Other academic groups have included histologic and genomic information in the testing algorithm. OBJECTIVE.­: To apply some of these triaging strategies for a more efficient identification of NTRK fusions within a single institution, so pathologists can gain practical insight on how to start looking for NTRK fusions. DESIGN.­: A multiparametric strategy combining histologic (secretory carcinomas of the breast and salivary gland; papillary thyroid carcinomas; infantile fibrosarcoma) and genomic (driver-negative non-small cell lung carcinomas, microsatellite instability-high colorectal adenocarcinomas, and wild-type gastrointestinal stromal tumors) triaging was put forward. RESULTS.­: Samples from 323 tumors were stained with the VENTANA pan-TRK EPR17341 Assay as a screening method. All positive IHC cases were simultaneously studied by 2 NGS tests, Oncomine Comprehensive Assay v3 and FoundationOne CDx. With this approach, the detection rate of NTRK fusions was 20 times higher (5.57%) by only screening 323 patients than the largest cohort in the literature (0.30%) comprising several hundred thousand patients. CONCLUSIONS.­: Based on our findings, we propose a multiparametric strategy (ie, "supervised tumor-agnostic approach") when pathologists start searching for NTRK fusions.

SIEVE: joint inference of single-nucleotide variants and cell phylogeny from single-cell DNA sequencing data.

We present SIEVE, a statistical method for the joint inference of somatic variants and cell phylogeny under the finite-sites assumption from single-cell DNA sequencing. SIEVE leverages raw read counts for all nucleotides and corrects the acquisition bias of branch lengths. In our simulations, SIEVE outperforms other methods in phylogenetic reconstruction and variant calling accuracy, especially in the inference of homozygous variants. Applying SIEVE to three datasets, one for triple-negative breast (TNBC), and two for colorectal cancer (CRC), we find that double mutant genotypes are rare in CRC but unexpectedly frequent in the TNBC samples.

Single-cell mtDNA heteroplasmy in colorectal cancer.

Human mitochondria can be genetically distinct within the same individual, a phenomenon known as heteroplasmy. In cancer, this phenomenon seems exacerbated, and most mitochondrial mutations seem to be heteroplasmic. How this genetic variation is arranged within and among normal and tumor cells is not well understood. To address this question, here we sequenced single-cell mitochondrial genomes from multiple normal and tumoral locations in four colorectal cancer patients. Our results suggest that single cells, both normal and tumoral, can carry various mitochondrial haplotypes. Remarkably, this intra-cell heteroplasmy can arise before tumor development and be maintained afterward in specific tumoral cell subpopulations. At least in the colorectal patients studied here, the somatic mutations in the single-cells do not seem to have a prominent role in tumorigenesis.