Identification of an optimal algorithm for effective diagnostics of non-small cell lung cancer with ALK gene rearrangement - implementation of the method and practical experiences with routine diagnostics
Abstract
The aim of retrospective (RS) part of the study was a) to find an immunohistochemical (IHC) detection procedure of ALK protein with specificity and sensitivity high enough to use this antibody as screening method for selecting non-small cell lung cancer (NSCLC) cases for fluorescence in situ hybridization (FISH) testing of ALK gene rearrangement and b) to determine diagnostic yield of endo- and trans- bronchial and transthoracic biopsies and cytoblocks for ALK gene rearrangement testing. The best IHC method of ALK protein detection (clone D5F3, dilution 1:100, Cell Signaling Technology, Danvers, MA, USA) was then verified in prospective (PS) routine testing of patients with NSCLC.
ALK status was correlated with tumor morphology and clinical data. In the RS part, 170 EGFR-nonmutated cases of NSCLC were IHC and FISH tested.
In the PS part, 557 cases of NSCLC were tested by IHC and 76 by FISH. There were 8/154 (5.2%) cases with ALK gene rearrangement detected in the RS part and 24/557(4.3 %) in the PS part.
Sensitivity and specificity of the best IHC method were 100 % and 99 % in the RS part and 100 % and 80 % in the PS part. Diagnostic yield of "small" biopsies was between 74 - 80 % retrospectively, depending on IHC variant, and 88 % prospectively.
No case with ALK gene rearrangement detected prospectively had EGFR mutation. High diagnostic yield confirms that ALK status testing can be used in this type of specimen.
Prevalence of 5.2 % in the RS part (EGFR-nonmutated cases) and 4.3 % in the PS part (without known EGFR mutation status), tumor morphology (solid and acinar type, mucinous type or at least partial mucin production) as well as lower average age and male/female ratio of patients with ALK positive tumors in the PS part (57.5 y vs. 65.2 y, 8 men/16 women vs. 336 men/197 women) are consistent with global data.