This article is mentioned in:

Introduction

Recent advances in the biology of non-small cell lung cancer (NSCLC) have identified driver mutations that can predict response to the therapy. This molecularly targeted therapy has dramatically improved results for patients whose tumors harbor somatically activated oncogenes, such as epidermal growth factor receptor (EGFR), anaplastic lymphoma receptor tyrosine kinase (ALK), ret proto-oncogene (RET) or ROS proto-oncogene 1 (ROS1) (1–3). Although EGFR mutations and ALK, RET and ROS1 translocations are predominantly detected in lung adenocarcinomas from never-smokers and light smokers, these associations are not absolute; thus, clinical profile must not be used to exclude patients from molecular testing for specific alterations (4).

Activating rearrangements of ALK gene are detectable in 1.9–6.8% of NSCLC cases (5–9). In these tumors, a fusion gene involving echinoderm microtubule associated protein-like 4 (EML4) and ALK (EML4-ALK) encodes a chimeric protein with constitutive tyrosine kinase activity that confers transformation potential and responsiveness to specific targeted inhibitors (5). Whereas the fused ALK gene region remains constant (exon 20), different EML4 exons can be affected, generating several variants; other ALK fusion partners have also been reported (10,11). In patients with ALK-positive NSCLC, the use of crizotinib, an oral ATP-competitive selective inhibitor of the ALK tyrosine kinase, along with a new generation of ALK tyrosine kinase inhibitors (alectinib and ceritinib), has demonstrated superiority over standard chemotherapy (3,4).

Fluorescence in situ hybridization (FISH), bright-field dual-color chromogenic in situ hybridization (12), immunohistochemistry (IHC) (7,12) and reverse transcriptase-polymerase chain reaction (13) are available methods for the detection of ALK rearrangements. ALK FISH has been recognized as the most sensitive method (9). Recently, the FDA has approved the Ventana ALK (D5F3) CDx Assay (Roche, Tucson, AZ, USA) to identify patients for crizotinib therapy. Ultrasensitive IHC assays may obviate FISH confirmation in positive IHC cases; however, the likelihood of false negative IHC results demands FISH testing, at least in some situations (7,8,12). Currently, the only approved FISH assay for the detection of ALK positivity, as a companion diagnostic tool for crizotinib-based treatment eligibility, is the Vysis ALK Break Apart FISH Probe Kit (Abbott Molecular, Des Plaines, IL, USA).

In the current study, we describe our experience of FISH ALK rearrangement detection in 2,045 routine diagnostic samples in the public health care system of Galicia (Spain). This report is of particular interest for comparative purposes due to Galicia's isolation in North-Western Spain and its specific cultural characteristics, which may contribute to the emergence of genetic diversity (14).

Materials and methods

Patients and samples

A total of 2,045 unselected samples of NSCLC from patients in 9 hospitals of the Galician public health care system (Spain), collected between October 2010 and July 2015, were investigated. Samples were received as buffered formalin-fixed, paraffin-embedded tissue biopsies, cell blocks and cytological smears. In the referral center (Clinical University Hospital, Santiago de Compostela, Spain), pathologists verified the adequacy of the samples and reviewed the diagnosis according to the World Health Organization criteria (1). For all patients, medical records were reviewed, and written informed consent obtained. The Independent Ethics Committee of Galicia (Galician Healthcare Service; SERGAS) approved the study protocol, which was conducted in accordance with the Declaration of Helsinki and applicable Spanish laws.

ALK FISH analysis

The ALK genetic status of each patient was evaluated by FISH in representative tumor areas, with an LSI ALK dual-color, break-apart rearrangement probe (Vysis ALK Break Apart FISH Probe Kit, Abbott Molecular Inc.), following the manufacturer's protocol. In each case, 50 non-overlapping nuclei were examined by fluorescence microscope (BX51; Olympus Corporation, Barcelona, Spain). The kit included red (3′) and green (5′) break-apart probes. ALK was determined to be not rearranged when the two signals were adjacent or fused (appearing yellow under an Orange/Green V2 filter) or when there was a single green signal. ALK was considered positive (rearranged) when ≥1 set of red and green signals were ≥2 signal diameters apart, or when a single red signal was present. A sample was considered negative for ALK if <5 cells out of 50 (<10%), and positive when >25 cells out of 50 (>50%), exhibited split ALK 5′ and 3′ probe signals or isolated 3′ signals. With borderline results, when the number of positive cells was 5–25 (10–50%), a second reader evaluated the slide: First and second cell count readings were added together and a percentage out of 100 cells was calculated. If the positive cell percentage was ≥15%, the sample was considered positive.

Statistical analysis

Statistical analysis was conducted with SPSS version 19.0 (IBM SPSS, Armonk, NY, USA). The association between ALK mutation status and clinicopathological features was calculated using Pearson's χ2 test or Fisher's exact test. P<0.05 was considered to indicated a statistically significant difference.

Results

ALK assessment

The 2,045 samples of NSCLC included 1,686 (82.4%) tissue biopsies, 306 (15%) cell block samples and 53 (2.6%) cytological smears. Among the total samples examined, 1,866 (91.2%) could be tested by FISH (Fig. 1). Specifically, 1,566 (92.9%) tissue biopsies, 256 (83.7%) cell blocks and 46 (86.8%) cytological smears could be assessed. In the remaining 179 (8.8%) samples, the FISH result was not obtained owing to insufficient/non-representative tumor material (n=143; 79.9%) or to assay failure (n=36; 20.1%) due to pre-analytical factors (poor fixation, decalcification treatment, etc.). The rate of successful evaluation was not significantly higher in histological samples compared with cytological samples (92.9% vs. 84.1%). Additionally, no differences were found between cell blocks and cytological smears (83.7% vs. 86.8%).

Figure 1. ALK break-apart FISH signals in non-small cell lung cancer. (A) ALK FISH-negative specimen. (B) ALK FISH-positive specimen showing split red and green signals. (C) ALK FISH-positive specimen showing isolated single red signals. ALK, anaplastic lymphoma receptor tyrosine kinase; FISH, fluorescence in situ hybridization.

ALK rearrangement

ALK rearrangements were identified in 82 (4%) out of 2,045 tumor samples: 65 (79.3%) tissue biopsies, 15 (18.3%) cell blocks and 2 (2.4%) cytological smears (Table I). The mean percentage of ALK-positive cells was 52% (range, 15–92%). A typical translocation pattern was observed in 62 (75.6%) cases, and the remaining cases (24.4%) exhibited a split pattern (Fig. 1).

Table I. Clinical and pathological features of patients positive and negative for ALK rearrangement.

Table I.

Clinical and pathological features of patients positive and negative for ALK rearrangement.

Variable	All cases	ALK-negative cases	ALK-positive cases	P-value
Total, n (%a)	2,045	1,784 (87.2)	82 (4.0)
Age, years				<0.001
Range	25–94	31–94	25–83
Median	64.0	64.4	58.0
Gender, n (%b)				<0.001
Male	1,420 (69.4)	1,262 (70.7)	31 (37.8)
Female	605 (29.6)	507 (28.4)	49 (59.8)
Unknown	20 (1.0)	15 (0.8)	2 (2.4)
Histology, n (%b)				0.017
Adenocarcinoma	1,559 (76.2)	1,352 (75.8)	72 (87.8)
Squamous cell carcinoma	263 (12.9)	239 (13.4)	1 (1.2)
Large cell neuroendocrine carcinoma	21 (1.0)	18 (1.0)	2 (2.4)
Adenosquamous carcinoma	17 (0.8)	16 (0.9)	0 (0.0)
Not otherwise specified	185 (9.0)	159 (8.9)	7 (8.5)
Smoking history, n (%b)				<0.001
Ex-smoker	783 (38.3)	688 (38.6)	20 (24.4)
Smoker	761 (37.2)	661 (37.0)	23 (28.0)
Non-smoker	372 (18.2)	312 (17.5)	34 (41.5)
Unknown	129 (6.3)	123 (6.9)	5 (6.1)
Stage, n (%b)				<0.001
T1 (T1a+T1b)	36 (1.8)	34 (1.9)	1 (1.2)
T2	39 (1.9)	38 (2.1)	1 (1.2)
T3	126 (6.2)	117 (6.6)	3 (3.7)
T4	883 (43.2)	780 (43.7)	47 (57.3)
Unknown	961 (47.0)	815 (45.7)	30 (36.6)

a Percentage of total patients

b percentage by ALK rearrangement group. ALK, anaplastic lymphoma receptor tyrosine kinase.

Clinicopathological findings

The clinicopathological features of the 2,045 NSCLC patients are summarized in Table I. The majority were male (69.4%) and ex-smokers (38.3%) or current smokers (37.2%). At diagnosis, the vast majority were in advanced stage (43.2% at stage IV). The predominant histological type was adenocarcinoma (76.2%). Significant associations between ALK rearrangement and younger age (P<0.001; Mann-Whitney U test), female gender (P<0.001; Pearson's χ2 test), non-smoking history (P<0.001; likelihood ratio test), stage IV at diagnosis (P<0.001; likelihood ratio test) and the adenocarcinoma histological type (P=0.046; likelihood ratio test) were found. Other ALK-positive tumor types were large cell neuroendocrine carcinomas (n=2), squamous carcinoma (n=1) and not otherwise specified carcinomas (n=7). ALK-positive adenocarcinomas (n=72) exhibited a predominantly solid pattern of growth in 45.8% of cases, with cellular features of signet ring cell carcinoma in 36.4% (n=12) of these (Table II and Fig. 2). No simultaneous EGFR mutations were detected in any cases.

Figure 2. Representative histological patterns of adenocarcinoma tissues. (A) Solid growth pattern, (B) solid pattern with a signet ring cell component, (C) acinar pattern, (D) cribriform pattern, (E) mucinous pattern, and (F) micropapillary pattern. Hematoxylin-eosin staining; original magnification, ×400.

Table II. Adenocarcinoma growth patterns of ALK gene rearrangement-positive patients (n=72).

Table II.

Adenocarcinoma growth patterns of ALK gene rearrangement-positive patients (n=72).

Pattern of growth	Cases, n
Solid adenocarcinoma NOS	19
Solid with signet ring cells	12
Acinar adenocarcinoma NOS	13
Cribriform adenocarcinoma	2
Colloid adenocarcinoma	4
Mixed adenocarcinomaa	5
Adenocarcinoma NOS	8
Unknown	9

a Including solid-papillary patterns (1 case), solid-cribriform patterns (1 case), acinar-cribriform patterns (1 case), acinar-solid patterns (1 case), and acinar-micropapillary patterns (1 case). NOS, not otherwise specified; ALK, anaplastic lymphoma receptor tyrosine kinase.

Discussion

To the best of our knowledge, this is the largest study on ALK rearrangements in unselected NSCLC patients in a Spanish Caucasian population. It is also the first study in Galicia, North-Western Spain, and is particularly interesting as the adequacy and histology of all samples were reviewed in a single institution by the same pathologists using the same criteria. Although Galicia is relatively isolated, the current results confirm the low frequency of ALK rearrangements (4%) in unselected Caucasian NSCLC patients, which is marginally higher than values from other areas of Spain (3%) (6), and within the range of 1.9–6.8% reported in the literature (5,7–9).

The current study utilized the Food and Drug Administration (FDA)-approved ALK break-apart FISH method, with excellent results. The successful evaluation rate difference between tissue samples and cytological samples was not significant (92.9% vs. 84.1%); very similar results were obtained using cell blocks and cytological smears (83.7% vs. 86.8%). Although technical limitations have been reported in cytological specimens (6), the success rate in the present study was in line with the literature, which reports the feasibility of ALK FISH analysis of cytological samples as ranging from 79–97% (15). Cytological smears may be preferable for FISH assays as they avoid formalin fixation problems and permit the evaluation of whole nuclei, eliminating overlapping nuclei and nuclear truncation (16). According to FDA guidelines (12), this FISH analysis is applicable on histological samples, however, cytological smears could also be appropriate.

The present study revealed an apparent association of ALK rearrangement with non-smoking, female gender, younger age, stage IV and adenocarcinoma. The significant proportion of ALK-positive non-smokers is consistent with previous reports (9), but 55.9% of the current cases were previous or current smokers, which confirms that smoking is not a reliable criterion on which to predict or exclude ALK rearrangement. ALK positivity was determined to be strongly associated with female gender (P<0.001); however, controversial results have been reported (6,8). A marginally, non-significantly higher rate of ALK positivity in females has been described in a recent meta-analysis of 6,950 patients (9). The results of the present study and that of Proietti et al (16) found a younger age to be associated with ALK rearrangement. We also found ALK positivity to be associated with stage IV disease; however, analysis of a greater number of patients in other stages would be desirable to reach reliable conclusions. Consistently with previous reports (6,8,9), the current results found ALK rearrangements to be largely, but not completely, restricted to adenocarcinomas, specifically solid or acinar variants or with signet ring cell features. ALK positivity was also identified in one squamous cell carcinoma and two large cell neuroendocrine carcinomas.

In conclusion, the present findings indicated that ALK evaluation by FISH is feasible in tissue and cytological samples. It was also found that ALK translocation is associated with a non-smoking history, a younger age, the female gender, stage IV and the adenocarcinoma histological type. Due to the fact that ALK rearrangement was detected in various histological tumor types, the ALK study should not be restricted to adenocarcinomas.

References