Aberrant anaplastic lymphoma kinase expression in high-grade pulmonary neuroendocrine carcinoma
- Harumi Nakamura1,
- Koji Tsuta1,
- Akihiko Yoshida1,
- Tatsuhiro Shibata2,
- Susumu Wakai1,
- Hisao Asamura3,
- Koh Furuta1,
- Hitoshi Tsuda1
- 1Department of Pathology and Clinical Laboratories, National Cancer Center Hospital, Tokyo, Japan
- 2Division of Cancer Genomics, National Cancer Center Research Institute, Tokyo, Japan
- 3Division of Thoracic Surgery, National Cancer Center Hospital, Tokyo, Japan
- Correspondence to Dr K Tsuta, Department of Pathology and Clinical Laboratories, National Cancer Center Hospital, 5-1-1 Tsukiji, Chuo-ku, Tokyo 104-0045, Japan;
- Received 2 November 2012
- Revised 15 March 2013
- Accepted 29 March 2013
- Published Online First 25 April 2013
Aims In lung cancer with anaplastic lymphoma kinase (ALK) gene rearrangement, accurate diagnosis is essential to identify patients who can be treated with a specific kinase inhibitor. Sensitive ALK immunostaining can provide nearly complete concordance with gene rearrangement status and is expected to serve as a surrogate biomarker for tailored treatment with an ALK inhibitor. In the present report, we describe aberrant ALK expression in a small proportion of pulmonary neuroendocrine carcinomas (NECs) that do not have ALK gene alteration.
Methods A total of 227 pulmonary NECs were assembled on tissue microarrays and were subjected to highly sensitive ALK staining methods.
Results We observed focal positivity with heterogeneous intensity in 2 (2.9%) of 69 small-cell carcinomas and 1 (0.9%) of 106 large-cell NECs. In contrast, 52 carcinoid tumours were all negative for ALK expression. Neither ALK rearrangement nor amplification was observed using fluorescence in situ hybridisation, and no somatic mutation was detected in three ALK positive NECs.
Conclusions Thus, this aberrant expression is probably of a wild-type ALK and a potential pitfall when implementing sensitive ALK immunohistochemistry in the molecular diagnosis of lung cancer.
The advent of therapy targeted to a driver mutation has dramatically enhanced the treatment options for lung cancer patients. Recent discovery of anaplastic lymphoma kinase (ALK) rearrangement in a small subset of lung cancers1 rapidly culminated in a clinical trial of crizotinib (an ALK/mesenchymal epithelial transition inhibitor), which showed successful disease control in many patients with this disease.2 Now, just 4 years after the discovery of ALK-rearranged lung cancer, Crizotinib has been approved by the US Food and Drug Administration and subsequently the Ministry of Health, Labor, and Welfare in Japan.
At present, fluorescence in situ hybridisation (FISH) and reverse transcription-PCR (RT-PCR) are the two most widely used molecular techniques to detect ALK rearrangement.1 However, these assays can be labour-intensive and costly as well as require expertise that may not be available in all laboratories. Immunohistochemistry has therefore attracted much attention as a potential surrogate for these molecular assays. ALK-rearranged lung cancer expresses a relatively modest level of ALK protein compared with anaplastic lymphoma, and thus the conventional ALK staining method that has been optimised for lymphoma diagnosis is not suitable for reliable screening.3 Efforts to develop highly sensitive ALK staining techniques have been successful in a number of institutions including ours by using high-affinity antibodies and/or an optimised signal amplification system.3–5 These sensitive ALK staining protocols have shown almost complete concordance with molecular tests by RT-PCR or FISH, and immunohistochemistry is now expected to function as a potential diagnostic tool that may substitute for other ALK molecular assays in the near future.
In our diagnostic service using sensitive ALK immunohistochemistry, we recently encountered a case of pulmonary large-cell neuroendocrine carcinoma (LCNEC) that labelled for ALK and yet lacked ALK rearrangement by FISH. Because our staining protocol was previously validated to achieve complete concordance with ALK rearrangement status when applied to lung adenocarcinoma,5 this rearrangement-independent ALK expression prompted us to undertake a systematic survey of ALK expression in lung neuroendocrine carcinomas (NECs).
Material and methods
The institutional review board approved the study protocol (2010-075). The specimens used in this study were from 227 patients who underwent surgical resection for NEC at the National Cancer Center Hospital (Tokyo, Japan) between 1982 and 2010. Histological diagnosis was based on the latest edition of WHO classification.6 The 227 NECs consisted of 52 carcinoids (including both typical and atypical types), 69 small-cell lung carcinomas (SCLCs) and 106 LCNECs.
The most representative tumour areas were sampled for tissue microarray (TMA) analysis. The TMAs were assembled with the tissue-arraying instrument KIN-1 (Azumaya, Tokyo, Japan). To reduce sampling bias due to tumour heterogeneity, we used duplicate core samples measuring 2.0 mm in diameter taken from two different areas of each tumour. All 227 NECs were immunohistochemically stained using the ALK antibody with TMA sections. We used three methods for ALK immunohistochemistry: (1) a conventional method using monoclonal mouse antihuman CD246, ALK protein, clone ALK1 (1 : 50; DAKO, Carpinteria, California, USA) that was optimised for lymphoma diagnosis; (2) highly sensitive methods that we had previously validated and used routinely at our laboratory involving clone 5A4 (1 : 40; Abcam, Cambridge, UK) and the Envison-FLEX plus system (DAKO);5 and (3) another highly sensitive method that was previously validated in another laboratory using antibody clone D5F3 (1 : 250; Cell Signaling Technology, Danvers, Massachusetts, USA3) and the manufacture's recommended detection system (SignalStain Boost IHC Detection Reagent; Cell Signaling Technology).
Among the ALK positive cases detected through TMA section to reveal the ALK staining pattern and indicate regions of tumour tissue, we performed the immunostaining for three ALK antibodies with whole sections as well.
Paraffin-embedded tissue sections (4 μm thick) were used for FISH analysis. To reveal the correlation between protein expression and gene alteration, we used the serial cut whole sections for immunostaining and FISH analyses.
To assess ALK rearrangement, we used LSI break-apart ALK (2p23) probes (Abbott Molecular Inc., Des Plaines, Illinois, USA) in accordance with the manufacturer's instructions. At least 50 non-overlapping tumour cells were counted, and cases with more than 15% cells showing split-apart signals or lone 3′ signals were considered positive for ALK rearrangement.
To assess ALK amplification, we used ALK/CEP2p probes (GSP Laboratory, Kawasaki, Japan) in accordance with the manufacturer's instructions. At least 50 non-overlapping tumour cells were examined and the positive result of copy number gain was defined as ALK/CEP2p ≥2.0.
Analysis of ALK mutational status
We extracted DNA from sections of methanol-fixed paraffin-embedded tumour specimens from ALK-immunopositive NECs. Exons 23 and 25 of the ALK gene, which are hotspot-activating mutations, were amplified by PCR using EX Taq HS polymerase (TAKARA, Shiga, Japan). The primers used were ALK-EX23F: 5′ GCCTTTATACATTGTAGCTGC 3′, ALK-EX23R: 5′ TCGGAGGAAGGACTTGAGGTC 3′, ALK-EX25F: 5′ TCTTCCCAGAGACATTGCTGC 3′ and ALK-EX25R: 5′ GGTAGAAAGTTGACAGGGTAC 3′. The PCR products were purified (QIAquick PCR purification kit; Qiagen, Valencia, California, USA) and analysed by sequencing with the same primers (Big Dye sequencing kit; Applied Biosystems, Carlsbad, California, USA).
ALK immunoreactivity was observed in 3 (1.3%) of the 227 NECs when using the highly sensitive immunohistochemical methods. Both the sensitive methods (ie, 5A4 with EnVision Flex Plus and D5F3 with SignalStain Boost) yielded similar results. Immunopositive cases comprised 2 (2.9%) of 69 SCLCs (figure 1A,B) and 1 (0.9%) of 106 LCNECs (figure 1C). No carcinoids were immunopositive. The pattern of staining was focal and heterogeneous in intensity, ranging from weak to strong within the same tumour, and generally cytoplasmic with some plasma membranous accentuation exhibiting a granular quality. The almost similar staining pattern and immunopositive region of all positive cases was observed with both clone 5A4 and D5F3 with both TMA and whole section. In contrast, no case was found to be immunopositive using the conventional ALK1 staining protocol.
FISH analysis showed no evidence of ALK rearrangement in any of the three ALK-immunopositive cases. In addition, no gene amplification was observed using ALK/CEP2p probes for any of these three immunopositive cases, and mutation analysis revealed no evidence of ALK mutations.
Because the clinical activity of ALK inhibitors is determined by ALK rearrangement status, it is important to accurately detect ALK rearrangement in tumours.2 The status of ALK expression detected by recently developed sensitive immunohistochemistry has almost perfect concordance with the status of ALK rearrangement detected by RT-PCR or FISH and is expected to be applicable to daily tumour diagnosis.3–5 In the present study, however, we showed that a small subset (3%) of high-grade NECs exhibited aberrant ALK reactivity unassociated with ALK gene alteration. Furthermore, since ALK positivity was evaluated only in a part of the tumour by using TMA sections, the true incidence of high-grade NECs with rearrangement-independent ALK immunopositivity may be more frequent. In contrast, none of the low-intermediate-grade NECs (ie, carcinoids) labelled for ALK.
The mechanism of ALK immunoreactivity in high-grade NECs remains unclear. The reactivity likely represents true protein expression rather than a technical artefact because it was observed in two independent staining methods by using different antibody clones and detection systems. Further, we used a biotin-free detection method and the reaction should therefore not reflect non-specific endogenous biotin staining. The three ALK-immunopositive cases had no ALK rearrangement, amplification or mutations. Other mechanisms such as epigenetic regulation or overstabilisation of the protein may result in immunopositivity. Takeuchi et al7 similarly noted that a small proportion of SCLCs and LCNECs were ALK positive using a sensitive staining method, and concluded that this aberrant expression was caused by a wild-type ALK. It is noteworthy that the ALK protein is normally expressed, albeit weakly, in the central nervous system,8 and it may be that neuroendocrine differentiation in SCLCs and LCNECs gives rise to aberrant ALK expression, although this does not explain the lack of reactivity in carcinoids.
There are currently no data on whether tumours with aberrant expression of ALK are sensitive to ALK inhibitors. However, because the effectiveness of inhibitors has been proven only in cases with ALK rearrangement, the aberrant ALK expression that we describe here should be differentiated from true rearrangement-associated ALK positivity. This differentiation should ultimately be achieved by molecular methods; however, we believe that there are a few histological features that help identify true rearrangement-associated expression. First, the pattern of ALK immunoexpression in high-grade NECs was focal with heterogeneous intensity. In contrast, the staining pattern of ALK rearrangement lung cancers was typically diffuse and cytoplasmic, and virtually all the cells in the cores were labelled.9 Second, ALK rearrangement in lung cancer is typically associated with adenocarcinoma,5 ,7 and it is rare in high-grade NECs.10 However, not many cases of high-grade NECs were systematically studied for this molecular abnormality. In addition, combined histology showing both adenocarcinoma and high-grade NEC do occasionally occur. It has been reported that 10%–28% of high-grade NECs show combined histologies,11 ,12 and these cases may carry genetic changes characteristic of adenocarcinomas. Detailed histological analysis and staining for napsin A13 and mucin may be helpful in identifying a combined adenocarcinoma element.
In conclusion, a small subset of high-grade NECs can focally label for ALK by sensitive staining. This immunopositivity was not the result of gene rearrangement, mutation or amplification, and therefore these NECs are not candidates for kinase inhibitor treatment under the current guidelines. They should therefore be distinguished from cancers with true rearrangement-associated ALK expression. This under-recognised phenomenon is a potential problem and should be considered when incorporating sensitive ALK staining into lung cancer diagnostics.
A small subset of high-grade neuroendocrine carcinomas (NECs) can focally label for anaplastic lymphoma kinase (ALK) by sensitive staining.
This immunopositivity was not the result of gene rearrangement, mutation or amplification. Therefore, these NECs are not candidates for kinase inhibitor treatment under the current guidelines.
This under-recognised phenomenon is a potential problem and should be considered when incorporating sensitive ALK staining into lung cancer diagnostics.
We thank Ms Karin Yokozawa, Sachiko Miura, MT, Chizu Kina, MT, and Ms Shouko Ohashi for technical assistance.
Contributors Study concepts and study design: HN, KT and AY. Data acquisition: TS, SW, HA, KF and HT. Quality control of data and algorithms: HN, KT, AY, TS and HT. Data analysis and interpretation: HN, KT, TS, AY, SW and KF. Statistical analysis: HN and KY. Manuscript preparation: HN, KT and AY. Manuscript editing: HN, KY, AY and HT. Manuscript review: HN, KT, AY, TS, SW, HA, KF and HT. Planning, conduct and reporting of the work described in the article: HN and KT. Contributor(s) as being responsible for the overall content as guarantor(s): KT. Conception and design: HN, KT and AY. Financial support and administrative support: KT and HT. Provision of study materials or patients: KF and HA. Manuscript writing and final approval of manuscript: All authors.
Funding This work was supported in part by the National Cancer Center Research and Development Fund (23-A-2), (23-A-11) and (23-A-35).
Competing interests None.
Patient consent Obtained.
Ethics approval Approval of the study was granted by the Institutional Review Board and National Cancer Center Research Center Hospital.
Provenance and peer review Not commissioned; externally peer reviewed.