Torsten O. Nielsen MD/PhD FRCPC

Pathology and Laboratory Medicine University of British Columbia


While immunohistochemistry has achieved an integral role in pathology diagnosis, this protein-based technique has limitations in analytical reproducibility, quantification and multiplexing. Many biological features relevant to diagnosis and treatment are better assessed at the level of RNA gene expression, including, for example, breast cancer risk profiling, and identification of fusion oncogenes that are pathognomonic for an increasingly-large number of tumor types (particularly sarcomas, hematopoietic neoplasms and head & neck tumors). Genome-wide techniques (microarrays and RNA-seq transcriptomics) are excellent for research but are inefficient as diagnostic tools for reasons of both turnaround time and expense; moreover, most such techniques perform poorly when applied to the fragmented, low quality RNA present in standard formalin-fixed, paraffin-embedded pathology blocks.

RT-PCR can overcome this but is also inefficient in the multiplex context when >10 genes need to be assayed to define a subtype, assign a risk profile or cover variant exon fusion sites. NanoString technology is a recently-developed method ideally suited for quantitative digital analysis of up to 800 transcripts in a clinical specimen, a “sweet spot” that answers clinical questions quickly and efficiently. Dr. Nielsen will review why the PAM50 gene profile was switched from RT-PCR to NanoString as part of its clinical development as the Prosigna assay. He will also describe how this technology, adapted for the identification of fusion oncogenes, has recently become western Canada’s primary molecular diagnostic for sarcomas, replacing FISH and obviating the need for NGS assays that slower and more expensive.

Going forward, this technology is also being adapted to diagnose head & neck tumors as well as cancers characterized by highly variant fusions or by gene amplifications.


While a critical clinical distinction between hormone receptor positive and negative breast cancers has been understood for at least three decades, it was the advent of gene expression profiling technologies which really highlighted the large biological differences among the Luminal A, Luminal B, HER2 and basal intrinsic molecular subtypes. Although surrogate immunohistochemistry panels can assign breast cancer subtype, challenges remain in perfecting their analytical reproducibility.

Analysis at the RNA level can overcome some of these challenges, for example by applying the PAM50 signature (developed by Nielsen & colleagues) using the NanoString platform as the Prosigna assay. Several clinical tests based on gene profiles now exist that stratify risk among luminal breast cancers, some of which have achieved regulatory approval for clinical use and are incorporated into international guidelines for patient management decisions. The advent of next generation sequencing technologies has brought additional detail, with the latest research bringing RNA profiling to the single cell level.

As for the DNA genomic level, copy number aberration and point mutational patterns can refine subtype and identify cases that may be amenable to specific therapies such as PARP- or checkpoint- inhibitors. By sequencing primary tumors, all the commonly – and most of the uncommonly – mutated driver genes in breast cancer are now known. Dr. Nielsen will draw on recently-published data from his own group and others to review the key features of the molecular classification of breast cancer, and present his perspective on the near term future for how new technologies (including profiling of metastases, and “liquid biopsies”) may impact our biologic understanding, means of diagnosis and best management of breast cancer.

back to Clinical Pathology Speakers


Philippine Society of Pathologists, Inc. — Project F Sharp

All Rights Reserved 2018