Can you trust your gene fusion-negative results?

By Jeff Mitchell on Tue Sep 15, 2015

Rare CRTC3-MAML2 fusion identified in NSCLC sample

Archer™ FusionPlex™ assays are purpose-built to identify known and novel fusions via targeted NGS, with knowledge of only a single partner being required. One major advantage of this is fewer false negatives compared to other fusion detection techniques. The more targets in a FusionPlex assay, the lower the likelihood for a false negative. The FusionPlex Solid Tumor Panel is a comprehensive fusion panel with over 56 major partners and 300+ minor partners identified.

We recently screened 267 pan-negative FFPE non-small cell lung cancer (NSCLC) samples for common DNA hotspots and driver mutations using the FusionPlex ALK, RET, ROS1 v2 Panel. Negative samples were then orthogonally analyzed using the FusionPlex Solid Tumor Panel to identify any potential fusions.

Total nucelic acid was extracted from individual curls using the Qiagen® AllPrep® kit, and RNA quality was assessed using the Archer PreSeq™ RNA QC Assay. Libraries were prepared from 50ng RNA using the FusionPlex Solid Tumor Panel, sequenced using an Illumina® NextSeq® instrument, and subsampled to a depth of ~1.5 million reads. Sequencing results were then measured using Archer Analysis v3.2.1. A cutoff of 5 unique fusion molecules (the standard setting for fusion calling in Archer Analysis) was applied, and fusions were called. Interestingly, unique CRTC3-MAML2 fusion fragments were detected in one sample previously thought to be fusion negative (Figure 1).

Figure 1 CRTC3-MAML2 Read Pile-up

Figure 1. CRTC3-MAML2 fusion detected by FusionPlex Solid Tumor Panel. Depiction of the CRTC3-MAML2 fusion annotation and the supporting deduplicated read depth. Additionally, a read pile up showing the numer of supporting fragments identifying the CRTC3-MAML2 fusion.
» Click image to enlarge.

The CRTC3-MAML2 fusion is relatively novel, first appearing in the literature in 2008 and limited to only three references in the COSMIC database. The fusion protein includes the N-terminus of CRTC3 (Figure 2), a member of the CREB-regulated transcription coactivator gene family (Figure 3). Furthermore, this fusion was previously reported to be specific to mucoepidermoid carcinoma (MEC) (1), the most common primary malignancy of the salivary gland (2).

CRTC3-MAML2 mRNA fusion schematics

Figure 2. CRTC3-MAML2 fusion schematics. A) mRNA trancript for CRTC3, including exons 1 through 15 and 5' untranslated regions adjacent to exons 1 and 15. B) mRNA transcipt for MAML2, including exons 2 through 6 and untranslated regions adjacent to exons 2 and 6. C) mRNA transcript for the CRTC3-MAML2. The last observed CRTC3 exon is exon 1 while the first observed MAML2 exon is exon 3.

CRTC3-MAML2 Figure 3 CREb-regulated transcription coactivator gene family

Figure 3. The conserved gene promotor element CRE (cAMP response element) is regulated by cyclic AMP (cAMP). Transcription is stimulated via phosphorylated CREB dimer (CRE-binding protein) binding to CRE (3). CRTC3 family transcription factors regulate CREB-dependent gene transcription in a phosphorylation-independent manner (4).

This fusion appears to be extremely rare in NSCLC, as it has not been previously reported in the literature. However, pathologists with whom we consulted about this MEC-specific fusion detected in an NSCLC sample shared anecdotal evidence of at least one other incidence of this fusion in NSCLC.

In total, 10 replicate libraries were created with this sample and the FusionPlex Solid Tumor Panel according to the previously described method. The mean number of unique reads for these libraries was nearly 80, far exceeding the standard 5-unique-read cutoff for fusion verification by Archer Analysis. This data indicates that the CRTC3-MAMl2 fusion, previously reported only in MEC, is in fact real (Figure 4).

Figure 4 Unique reads supporting the CRCT3-MAML2 fusion

Figure 4. The number of unique reads supporting the CRCT3-MAML2 fusion exceeds the standard cutoff of 5 for each of the 10 library replicates generated.

This particular fusion would likely not have been detected if it were not screened with a comprehensive NGS assay like the FusionPlex Solid Tumor Panel. From a research perspective, this finding illustrates the importance of using broad-based approaches to screen for variants. Simply put, one would run out of tissue prior to identifying this mutation using sequential FISH, RT-PCR or probe-based capture methods.

Additionally, this finding brings into question the enpoint at which a sample can truly be called pan-negative. It seems that the current status quo in solid tumor analysis is to perform hotspot analysis of ~50 genes along with FISH assays for ALK, HER2 and MET. If the samples return no positive results, there are few analytical methods save clinical exome or genome sequencing to confirm that the sample is indeed fusion-negative. FusionPlex assays provide an opportunity to greatly increase the scope of fusion detection that can be done in-house, potentially identifying diagnostic or prognostic biomarkers that would otherwise go unnoticed.

Given the novelty of the fusion and its limited reporting outside of MEC, future detection in NSCLC and other cancer types will likely require similar tumor profiling.


  1. T. Nakayama et al., Clinicopathological significance of the CRTC3-MAML2 fusion transcript in mucoepidermoid carcinoma. Mod. Pathol. 22, 1575–1581 (2009).
  2. Oral Cancer Foundation, Mucoepidermoid Carcinoma. Accesses Sept 21, 2015.
  3. P. G. Quinn, D. K. Granner, Cyclic AMP-dependent protein kinase regulates transcription of the phosphoenolpyruvate carboxykinase gene but not binding of nuclear factors to the cyclic AMP regulatory element. Mol. Cell. Biol. 10, 3357–3364 (1990).
  4. NCBI, CRTC3 CREB regulated transcription coactivator 3 [ Homo sapiens (human) ]. Accessed Sept 21, 2015.

For Research Use Only. Not for use in diagnostic procedures. For Research Use Only. Not for use in diagnostic procedures.

About Jeff Mitchell

Jeff received his MBA from Boston University with dual concentration in finance and health sector management. He holds a B.S. in Biochemistry from Colorado State University. His research experience includes time at both the University of Colorado at Denver and Merrimack Pharmaceuticals. Prior to joining the ArcherDX team, he worked as a consultant for a private firm focused on the IVD industry and at the Innovation department within Partners Healthcare. In his spare time he daydreams about being a world-class adventurer that uncovers ancient relics that contain mystical powers.

How to contact us


2477 55th Street, Suite 202

Boulder, CO 80301


Phone: (877) 771 1093

Phone: (303) 357 9001

All content © 2017 ArcherDX, Inc.

For Research Use Only. Not for use in diagnostic procedures. For Research Use Only. Not for use in diagnostic procedures.