Article Text
Abstract
Background A subset of lung cancers harbours the fusion gene echinoderm microtubule-associated protein-like-4–anaplastic lymphoma kinase (EML4-ALK). Recently, immunohistochemistry for ALK has shown sensitivity for the detection of EML4-ALK-positive lung adenocarcinoma almost equal to that of the fluorescence in situ hybridisation (FISH) assay.
Aims To study the clinicopathological features of EML4-ALK-positive lung adenocarcinoma in a large number of surgically resected samples using immunohistochemistry, in order to establish a useful screening method for EML4-ALK-positive lung adenocarcinoma.
Methods Immunohistochemistry for ALK was used to screen for EML4-ALK-positive lung adenocarcinomas in 254 cases of surgically resected samples.
Results EML4-ALK-positive cases were detected in 3.1% of lung adenocarcinomas (8/254). EML4-ALK-positive lung adenocarcinomas showed significant associations with intra- and/or extra-cytoplasmic mucin (p=0.0001), and cribriform pattern with excessive extracytoplasmic mucin (p<0.0001). Signet-ring cell appearance alone lacked significance (p=0.149).
Conclusion EML4-ALK-positive lung adenocarcinoma has a tendency to express a characteristic morphological pattern. The combined use of morphological feature analysis and immunohistochemistry may be a useful and cost effective screening method for EML4-ALK lung adenocarcinoma.
- Lung cancer
- EML4-ALK
- immunohistochemistory
- morphology
- mucin
Statistics from Altmetric.com
Introduction
Lung cancer is the major cause of cancer-related deaths in the world. It is known that some gene mutations are involved in carcinogenesis of the lung. Epidermal growth factor receptor (EGFR) is an example of a gene mutation in lung cancer. Activation of the mutation in EGFR defines a group of patients with sensitivity to the chemical inhibitor for the kinase activity of EGFR, accounting for about 10% of primary lung cancers.1 2 In 2007, Soda et al identified the fusion oncogene joining the echinoderm microtubule-associated protein-like-4 (EML4) and the anaplastic lymphoma kinase gene (ALK) in non-small-cell lung cancer (NSCLC).3 The detection of EML4-ALK-positive lung cancer is needed to identify lung cancer patients for molecular target therapy. Some authors recommend molecular cascade screening for the detection of EML4-ALK-positive lung cancer.4 The cascade is a stepwise approach to test for gene mutations in lung adenocarcinoma: first for KRAS, second for EGFR, and third for EML4-ALK translocation. However, it is an expensive and time-consuming method. Moreover, EML4-ALK-positive lung cancer accounts for only about 5% of all NSCLC patients according to previous reports.3 5–8 To find a more efficient screening method for EML4-ALK-positive lung cancer, more precise clinical features and histopathological findings of EML4-ALK-positive lung cancer must be determined.
A previous report showed that EML4-ALK-positive lung cancers are characterised by smaller tumour size, and found in younger patients who are non-smokers or light smokers.6 Another study reported that EML4-ALK-positive lung cancer showed acinar morphology9 and signet-ring cell appearance.5 10
Recently, immunohistochemistry for ALK has shown sensitivity for the detection of EML4-ALK-positive lung adenocarcinoma almost equal to that of the fluorescence in situ hybridisation (FISH) assay.11 In this study, the clinicopathological features of EML4-ALK-positive lung adenocarcinoma were studied in a large number of surgically resected samples using immunohistochemistry, in order to establish a useful screening method for EML4-ALK-positive lung adenocarcinoma.
Materials and methods
Patients and sample collection
This study included samples from 254 Japanese NSCLC patients with a diagnosis of lung adenocarcinoma who underwent surgery at Osaka Police Hospital (Osaka, Japan) between January 2000 and December 2009. Clinical data were collected from inpatient and outpatient medical records.
Histological analysis
All surgically resected lung tumour specimens were embedded in paraffin and serial 5 μm thick sections were prepared. The pathological examination was based on standard H&E stained slides from all blocks of tissues. Pathological stage and differentiation were evaluated according to the current international tumour node metastasis (TNM) staging system and the World Health Organization classification. All tumour slides (consisting in each case of 1–25 slides, with a mean of six slides) were reviewed, and classified according to WHO histological types. In this study, bronchioloalveolar carcinoma is a pure type of carcinoma without invasion. The mixed subtypes category contained varied lesions. Therefore, in this study, an additional predominance classification was used, as described in a previous report.6 Regardless of percentage, the most dominant subtype was regarded as the predominant subtype. The remaining subtypes were regarded as minor components.
Three pathologists (RJ, TY, MT) reviewed all tumour slides of all cases. The decision on discrepant cases was made by consensus when two of the three pathologists agreed. All discrepant cases (10 cases) resulted in consensus.
Immunohistochemical analysis
Unstained paraffin-embedded sections were deparaffinised in xylene, hydrated through and rinsed in distilled water. Heat-induced epitope retrieval was performed with EnVision FLEX Target Retrieval Solution, High pH (Dako, Carpinteria, California, USA). Slides were then incubated at room temperature with antibody to ALK (clone 5A4, 1:100, Abcam) for 30 min. To increase the sensitivity of detection, the intercalated antibody-enhanced polymer (iAEP) method was used with minor modifications.12 Slides were incubated at room temperature with EnVision FLEX+ Mouse Linker (Dako) for 15 min. The immune complexes were then detected with the dextran polymer reagent.
FISH assay
FISH was performed on formalin-fixed paraffin-embedded tumour tissues with use of fluorescently-labelled bacterial artificial chromosome clone probes specific to the ALK and the EML4 loci (EML4 RP11-996L7; ALK RP11-984I21, RP11-62B19) using a histology FISH accessory kit (Dako).
DNA extraction and mutation analysis of EGFR, KRAS
Because no fresh sample was available for this study, DNA was extracted from formalin-fixed paraffin-embedded tissues; mutations in the EGFR (exons 18–21) genes were analysed by the peptide nucleic acid locked nucleic acid PCR-clamp (PNA LNA PCR-clamp) technique13 and in the KRAS (codons 12,13) genes by direct sequencing.14 Mutation analysis was performed on 67 cases, for which informed consent was obtained from patients. The ethics committee of our institute approved the genetic analyses in the present study.
Statistical analysis
Statistical analysis for the tumour size and age was carried out using Student's t test. The values are shown as mean ± SD. The relationship between EML4-ALK expression and clinicopathological variables was analysed with the χ2 test or Fisher's exact test. Statistical significance was defined as p<0.05.
Results
Clinical presentation
Using immunohistochemistry for ALK, the expression of ALK was studied in 254 lung adenocarcinomas. Eight of 254 cases showed homogeneous ALK protein in the cytoplasm without nuclear staining. The percentage of positive tumour cells ranged from 80% to 100%. Varied intensity of ALK expression was observed. The FISH assay confirmed the presence of the EML4-ALK fusion gene in the eight cases that showed the ALK protein (figure 1).
Table 1 summarises details of the clinicopathological features of all cases. All patients were Japanese, ranging in age from 42 to 87 years, with a mean age of 66.4 years. The maximum diameter of their tumours ranged from 0.5 to 7.5 cm. EML4-ALK-positive lung adenocarcinomas were significantly less differentiated than the EML4-ALK-negative lung adenocarcinomas (p=0.0025). EML4-ALK-positive and EML4-ALK-negative lung adenocarcinomas showed no difference in age, sex, tumour size, smoking habit and TNM stage. However, EML4-ALK-positive lung adenocarcinomas showed a tendency for association with younger patients, positive lymph nodes, adverse pathological stage and smoking habit. All of the EML4-ALK-positive cases had no mutation of the EGFR and KRAS genes.
Pathological findings and immunohistochemistry
Histologically, the 254 adenocarcinomas were comprised of 153 mixed subtypes, 34 acinar, 30 papillary, 2 others and 35 bronchioloalveolar carcinomas based on the WHO classification. The mixed subtypes category contained varied lesions. Therefore an additional predominance classification was used, as described in a previous report.6 According to the predominance subtyping, the 254 adenocarcinomas were comprised of 85 acinar, 83 papillary, 74 bronchioloalveolar non-mucinous carcinomas, 7 bronchioloalveolar mucinous carcinomas and 5 others (table 2).
Table 2 also summarises the histological features of all cases. According to the predominance subtypes, six of eight EML4-ALK-positive lung adenocarcinomas (75%) were subclassified as acinar-predominant adenocarcinomas and the other two cases were papillary-predominant adenocarcinomas. When compared with EML4-ALK-negative lung adenocarcinoma, the association of EML4-ALK-positive lung adenocarcinoma and acinar morphology was statistically significant (p=0.0184, Fisher's exact test). In other words, six of 85 (7.0%) acinar-predominant adenocarcinomas were positive for EML4-ALK fusion.
Lung adenocarcinoma cells have intracytoplasmic mucin in varying degrees and sometimes have extracytoplasmic mucin. Some lung adenocarcinoma cells show signet-ring cell like appearance because of excessive intracytoplasmic mucin (figure 2). Also, other lung adenocarcinomas show a cribriform pattern with excessive extracytoplasmic mucin (figure 3). In the 254 cases examined, 36 had intracytoplasmic or extracytoplasmic mucin detected by Alcian blue stain.
EML4-ALK-positive lung adenocarcinomas showed significant associations with intra- and/or extra-cytoplasmic mucin (p=0.0001), and cribriform pattern with excessive extracytoplasmic mucin (p<0.0001). Signet-ring cell appearance alone lacked significance (p=0.149) (table 3).
Two of eight EML4-ALK-positive cases did not have a characteristic morphological feature, nor did they show any mucin production (tables 4 and 5).
Discussion
A large-scale screening conducted for EML4-ALK-positive lung adenocarcinomas in our institute detected eight cases. In this study, the association of EML4-ALK-positive lung adenocarcinoma and acinar-predominant morphology, extra- and/or intra-cytoplasmic mucin production and cribriform pattern with excessive extracytoplasmic mucin was demonstrated. A previous report showed that EML4-ALK-positive lung cancers were characterised by smaller tumour size, younger patients and non-smokers or light smokers when compared against EML4-ALK-negative lung cancers.5 6 10 15 In this study EML4-ALK-positive lung adenocarcinoma was less differentiated, as described in a previous report.6 However, EML4-ALK-positive and -negative lung adenocarcinomas showed no difference in age, sex, tumour size, smoking habit and TNM stage. The screening was limited in this study to lung adenocarcinomas of Japanese patients. Previous reports showed that lung cancers harbouring the EML4-ALK fusion gene were mostly adenocarcinomas with a few exceptions.10 15 All EML4-ALK-positive lung adenocarcinomas had no mutation in the EGFR and KRAS genes, which concurred with previous reports.6 9 15
A previous report revealed the association of EML4-ALK fusion and acinar-predominant morphology9 and signet-ring cell appearance.5 10 In this study, acinar-predominant morphology and cribriform pattern with excessive extracytoplasmic mucin were useful factors, suggesting the presence of the EML4-ALK fusion gene, but signet-ring cell appearance by itself was not a useful factor. Previous reports revealed an association of EML4-ALK fusion and signet-ring cell appearance. However, the subjects screened were in a Western population or with limited clinical characteristics.5 10 There may be an underlying difference in the subject population by race and clinical characteristics. A recent report showed that primary signet-ring cell carcinoma of the lung shares many clinicopathologic characteristics with EML4-ALK-positive lung cancers.16 In point of fact, one case of EML4-ALK-positive lung adenocarcinoma in our study had signet-ring cell appearance. However, because the population of tumour cells appearing like signet-ring cells was very low, it was not a useful independent factor, suggesting the presence of the EML4-ALK fusion gene. It may however, be a clue in suggesting the presence of the EML4-ALK fusion gene because of excessive mucin production.
Previous reports have shown the utility of immunohistochemistry for EML4-ALK.11 12 Our study showed the usefulness of characteristic morphological features in screening for EML4-ALK-positive lung adenocarcinomas. The combined study of typical morphological features of EML4-ALK-positive lung adenocarcinomas by light microscopy followed by immunohistochemical confirmation may enable the screening for EML4-ALK-positive lung adenocarcinomas to be simple and cost effective.
Although light microscopic observations of specific morphological features and immunohistochemistry are very useful, two of our eight EML4-ALK-positive cases did not have characteristic morphological features. The two cases showed acinar-predominant morphology but no mucin production (figure 4 and table 5). A large-scale, multi-institute study will be needed to offer more insight into EML4-ALK-positive lung adenocarcinomas.
In this study, the FISH assay was performed only on positive cases. Standard methods are not yet established for detection of the EML4-ALK fusion gene in lung cancers. As previous reports showed,5 17 both FISH and immunohistochemistry may have false negative results. However, to perform the FISH assay on all screening cases at any institute is a costly undertaking. Immunohistochemistry, on the other hand, has been used widely in the diagnostic laboratory. It is easy to perform and relatively inexpensive. Recently, immunohistochemistry for ALK has shown almost equal sensitivity to that of the FISH assay for the detection of EML4-ALK-positive lung adenocarcinomas.11 In addition, the ALK fusion genes are recently discovered oncogenes in lung cancers. Conceivably, there is an unknown fusion partner of ALK, which cannot be detected by the FISH assay. Therefore, immunohistochemistry appears to be an appropriate screening method in the diagnostic laboratory.
In conclusion, we showed that EML4-ALK-positive lung adenocarcinomas had a tendency for expressing a characteristic morphological pattern. Continued investigations await the validation of the results presented in this study with respect to the potential usefulness of morphological analysis and immunohistochemistry in screening for EML4-ALK lung adenocarcinomas.
Take-home messages
EML4-ALK-positive lung adenocarcinomas were significantly less differentiated than the EML4-ALK-negative lung adenocarcinomas.
EML4-ALK-positive lung adenocarcinomas showed significant associations with intra- and/or extra-cytoplasmic mucin, and cribriform pattern with excessive extracytoplasmic mucin.
The combined study of typical morphological features of EML4-ALK-positive lung adenocarcinomas by light microscopy followed by immunohistochemical confirmation may enable the screening for EML4-ALK-positive lung adenocarcinomas to be simple and cost effective.
Acknowledgments
The authors thank Drs Eiichi Morii and Ayumi Furumoto for helpful advice and discussion, and Mr Michihiko Noguchi, Ms Ayumi Tsukiyama, Mr Yasuhiro Hashimoto, Ms Kumi Tanaka and Ms Saori Fukuda for technical assistance.
References
Footnotes
Competing interests None.
Ethics approval This study was conducted with the approval of the ethics committee of Osaka Police Hospital.
Provenance and peer review Not commissioned; externally peer reviewed.