The reported prevalence of ALK rearrangement in NSCLC ranges from 2%-7%. The primary standard diagnostic method is fluorescence in situ hybridization (FISH). Recently, immunohistochemistry (IHC) has also proven to be a reproducible and sensitive technique. Reverse transcriptase-polymerase chain reaction (RT-PCR) has also been advocated and most recently the advent of targeted Next-Generation Sequencing (NGS) for ALK and other fusions has become possible. This study compares ALK evaluation with all 4 techniques in resected NSCLC from the large ETOP Lungscape cohort.
96 cases from the ETOP Lungscape iBiobank, with any ALK immunoreactivity were examined by FISH, central RT-PCR and NGS. H-score>120 defines IHC-positivity. RNA was extracted from the same formalin-fixed, paraffin-embedded tissues. For RT-PCR, primers covered the most frequent ALK translocations. For NGS, the Oncomine™ Solid Tumour Fusion Transcript Kit was used. The concordance was assessed using the Cohen's kappa coefficient (two-sided alpha≤5%).
NGS provided results for 77 out of the 95 cases tested (81.1%), while RT-PCR for 77 out of 96 (80.2%). Concordance occurred in 55 cases out of the 60 cases tested with all 4 methods (43 ALK-negative, 12 ALK-positive). Using ALK co-positivity for IHC and FISH as gold standard, we derive sensitivity for RT-PCR/NGS 70.0%/85.0%, with specificity 87.1%/79.0%. Combining either RT-PCR or NGS with IHC, the sensitivity remains the same, while the specificity increases to 88.7% and 83.9% respectively.
NGS evaluation with the Oncomine™ Solid Tumour Fusion transcript kit and RT-PCR proves to have high sensitivity and specificity advocating their use in routine practice. For maximal sensitivity and specificity, ALK status should be assessed using two techniques and a third one in discordant cases. We therefore propose a customizable testing algorithm. These findings significantly influence existing testing paradigms and have clear clinical and economic impact.