OBJECTIVEThe aim of this study was to establish a standardized protocol for detection of ALK protein expression and gene fusion in cytologic specimens.

METHODSLung adenocarcinoma cytologic specimens were collected from seven hospitals in Beijing city. A detection protocol for ALK protein expression and gene fusion was designed according to the results of comparative experiment. Ventana immunohistochemical (IHC) ALK(D5F3) detecting ALK protein expression was performed in 203 prepared formalin-fixed paraffin-embedded (FFPE) cell blocks. ALK gene fusion in 98 EGFR gene wild type cytologic specimens and in 4 bronchoalveolar lavage fluid (BL) samples was detected by quantitative reverse transcription polymerase chain reaction (qRT-PCR). ALK gene fusion in the Ventana IHC ALK (D5F3) positive samples was further tested by fluorescence in situ hybridization (FISH). Six patients with ALK IHC-positive result were followed up to analyze the responses of crizotinib therapy. Comparative experiments: (1) Comparison of the results of 4% neutral buffered formalin fixed for different time (24 h, 48 h, 72 h) on the Ventana IHC ALK (D5F3) staining was conducted in two cases of IHC ALK positive FFPE cell blocks; (2) Comparing qRT-PCR results for ALK fusion in samples from FFPE cell blocks and cytospin prepared slides in 10 cases of lung adenocarcinoma cytologic specimens.

RESULTSAmong the specimens examined using the standardized protocol recommended by this study, 229 cases of cytologic specimens met the diagnostic criteria of lung adenocarcinoma. Among them, 207 cases obtained ALK gene test results (by at least one method), with an ALK test ratio of 90.4% (207/229). FFPE cell blocks were successfully prepared in 203 cases, Ventana IHC ALK (D5F3) were successfully performed in all the 203 FFPE cell blocks (100%), and the ALK protein positive detection rate was 10.3% (21/203). ALK fusion was tested in 98 FFPE cytologic samples of EGFR wild types by qRT-PCR, and 96 out of 98 (97.96%) cytologic samples were successfully performed.18 out of 19 IHC ALK-positive cases were verified to be of ALK fusion status by qRT-PCR. The concordance rate was 94.7% (Kappa=0.967, P<0.001) between Ventana IHC ALK (D5F3) and qRT-PCR, and the sensitivity of the Ventana IHC ALK (D5F3) assay compared with qRT-PCR was 100% and the specificity was 98.7%. FISH assay was used to verify the positive cases detected by Ventana IHC ALK (D5F3) staining. Two cases of low tumor cell content FFPE samples obtained indefinite results by FISH test. The six patients with positive ALK protein expression received crizotinib therapy, and 5 paitents got treated effectively. For two ALK IHC positive cases, which were 4% neutral buffered formalin fixed for 72 h, the result of Ventana IHC ALK (D5F3) staining became weakened obviously and uneven. In 10 cases of samples, total RNA was extracted from FFPE cytologic sections and cytospin prepared slides, and the results of qRT-PCR test and ALK gene fusion showed good concordance.

CONCLUSIONSThe standardized protocol recommended in this study expands the detection types and quantity of cytologic specimens for ALK protein expression and gene fusion and increased the detection rate. Ventana IHC ALK (D5F3) is a reliable method for detecting ALK protein expression in FFPE cell blocks. The pathologic quality control procedure prior to Ventana IHC ALK (D5F3) is crucial for the accuracy of testing the ALK gene status. When FFPE cell blocks could not be prepared or prepared unsuccessfully from the cytologic specimens, qRT-PCR may be an alternative option for the detection of ALK gene fusion.