Prospective definition of a diagnostic algorithm for EML4-ALK gene fusions or surrogates in NSCLC in the routine diagnostic setting: The effect of several testing methodologies and sample types.


Session type:

Perry Maxwell1,2, Claire McGready2, Stephen McQuaid1,2, Neil Anderson1, Jacqueline James2, Manuel Salto Tellez2
1NI MPL, Belfast Health & Social Care Trust, Belfast, UK, 2NI MPL, Queen's University Belfast, Belfast, UK


The presence of the Echinoderm Microtubule-associated proteinlike 4 (EML4) and Anaplastic Lymphoma Kinase (ALK) gene fusion(s) in lung cancer indicate a likely response to kinase inhibitor therapy1,2. We aim to prospectively test a diagnostic algorithm for the assessment of patient suitability for ALK-inhibitor therapy in non-small cell lung cancer including cytology, small biopsies and resection specimens3.


Cases were analysed using Formalin Fixed Paraffin Embedded (FFPE) sections over a 3 month period including biopsies (22), resections (22) and cell blocks (7). All samples were analysed by IHC (Ventana), FISH (Vysis) and RT-QPCR (Amoy).


Of 85 cases received, 51 cases were suitable for EML4-ALK testing, of which 31 were female with an average age of 65 years. There were 22 adenocarcinoma, 19 squamous cell carcinoma and 10 cases of indeterminate type.

  • a) On IHC, macrophages consistently stained positively but were distinguishable from malignant cells. In 5/19 squamous cell carcinomas equivocal IHC was seen but were negative by FISH and RT-QPCR.
  • b) We found 3 cases positive for EML4-ALK by IHC, FISH and typed to the EML4 variant 5' (A or B) or exon 17, EAM8 variant) by RT-QPCR.
  • c) IHC was technically successful in all cases. FISH was technically suboptimal in 1 case (1.9%) and required repeat slides in 8/51 cases (15.7%), 4 of which were resection specimens.


We present an algorithm for adequate and sequential use of all three diagnostic modalities, taking into account the apparent shortcomings of each technology, namely a) relative difficulty in IHC interpretation; b) technical challenge of FISH testing in all samples and c) relative incomplete coverage of RT-PCR designs. The algorithm involves the use of IHC as a screening tool and, when positive or equivocal, FISH testing followed by RT-QPCR when FISH is equivocal or technically suboptimal.