REVIEW article

Front. Oncol., 16 March 2012

Sec. Cancer Molecular Targets and Therapeutics

Volume 2 - 2012 | https://doi.org/10.3389/fonc.2012.00024

A Screening Method for the ALK Fusion Gene in NSCLC

  • YM

    Yoshiko Murakami 1

  • TM

    Tetsuya Mitsudomi 2

  • YY

    Yasushi Yatabe 1*

  • 1. Department of Pathology and Molecular Diagnostics, Aichi Cancer Center Nagoya, Japan

  • 2. Department of Thoracic Surgery, Aichi Cancer Center Hospital Nagoya, Japan

Abstract

Lung cancer research has recently made significant progress in understanding the molecular pathogenesis of lung cancer and in developing treatments for it. Such achievements are directly utilized in clinical practice. Indeed, the echinoderm microtubule-associated protein-like 4–anaplastic lymphoma kinase (ALK) fusion gene was first described in non-small cell lung cancer in 2007, and a molecularly targeted drug against the fusion was approved in 2011. However, lung cancer with the ALK fusion constitutes only a small fraction of lung cancers; therefore, efficient patient selection is crucial for successful treatment using the ALK inhibitor. Currently, RT-PCR, fluorescent in situ hybridization (FISH), and immunohistochemistry are commonly used to detect the ALK fusion. Although FISH is currently the gold standard technique, there are no perfect methods for detecting these genetic alterations. In this article, we discuss the advantages and disadvantages of each method and the possible criteria for selecting patients who are more likely to have the ALK fusion. If we can successfully screen patients, then ALK inhibitor treatment will be the best example of personalized therapy in terms of selecting patients with an uncommon genotype from a larger group with the same tumor phenotype. In other words, the personalized therapy may offer a new challenge for current clinical oncology.

Introduction

Lung cancer research has recently made significant progress in understanding the molecular pathogenesis of lung cancer and in developing treatments for it. For example, epidermal growth factor receptor (EGFR) tyrosine kinase inhibitors (EGFR–TKIs) have been developed, and a subset of lung cancer patients have shown dramatic therapeutic responses to these treatments. Subsequently, this superior response was revealed to be associated with the EGFR mutation, which was identified in 2005 (Lynch et al., ; Paez et al., ). Currently, EGFR–TKIs are the first-line treatments for patients with advanced disease when the tumor is positive for the EGFR mutation. More recently, a similarly marked response to an anaplastic lymphoma kinase (ALK) inhibitor was demonstrated in patients with ALK fusion-bearing lung cancer (Kwak et al., ).

The echinoderm microtubule-associated protein-like 4 (EML4)–ALK fusion in non-small cell lung cancer (NSCLC) was first described in 2007 (Rikova et al., ; Soda et al., ). This fusion results from a small inversion within chromosome 2p that leads to the expression of a chimeric tyrosine kinase in which the N-terminal half of the EML4 is fused to the intracellular ALK kinase domain. The strong in vitro (Soda et al., ) and in vivo (Soda et al., ) oncogenic activity of the fusion protein has been demonstrated, and the inhibition of the fused molecule with ALK inhibitors leads to a marked deterioration of the tumor cells in vitro, supporting that ALK fusion is a driver mutation of lung cancer. A dramatic response to the ALK inhibitor similar to that of the EGFR–TKIs was expected and was observed. However, the prevalence of the fusion is low, ranging from 2 to 4% in the literature (Inamura et al., ; Koivunen et al., ; Perner et al., ; Shinmura et al., ; Takahashi et al., ) and in our consecutive series (Table 1). Even when the patients are selected for their predictive clinical characteristics, such as a lower median age and not having a history of smoking, it is difficult to identify the subsets of ALK-positive tumors. Therefore, efficient screening for the ALK fusion gene is a crucial issue in clinical practice.

Table 1

CharacteristicsAll patients (n = 811)Genotype
ALK (n = 31)EGFR (n = 324)KRAS (n = 62)Null (n = 331)
AGE, YEAR
Median63561656465
Range22–8935–7631–8941–8022–83
SEX
Female36221219815111
Male4491012647210
SMOKING STATUS
Never smoked3592132111499
Smoker4521011248222
HISTOLOGY
Adenocarcinoma7133032256231
Squamous cell5210248
Adenosquamous80117
Other3800335
PATHOLOGICAL STAGE
I4921822540168
II1374341072
III150854970
IV32110311

Clinicopathological characteristics of ALK-positive lung cancers.

1Statistically significant by the independent sample two-tailed t-test (p < 0.002).

2Statistically significant difference in sex between the ALK-positive and negative tumors (p = 0.001).

3Statistically significant difference in smoking status between the ALK-positive and negative tumors (p = 0.001).

Approximately one-third of lung cancer patients are diagnosed with localized disease that can be treated by surgical resection (Molina et al., ); however, nearly 30–35% will relapse after the initial surgery, even following a diagnosis at the earliest stage. For these patients, the surgically resected tissue may serve as an important sample for further molecular study because the recurrent tumor is often impossible to access without invasive procedures. In contrast, the other two-thirds of patients present with advanced, inoperable disease, and therefore a biopsy or a cytology specimen may be the only available tissue. Therefore, the methods used for detecting the ALK fusion should be applicable to these small tissue and cytology specimens. Currently, there are three methods for ALK testing: RT-PCR, fluorescent in situ hybridization (FISH) and immunohistochemistry. Because there is no perfect method, we need to know the applicable range of the samples and the sensitivity, specificity and limitations of the individual methods. Before explaining the details of these methods, we will consider whether the targeted patient population can be identified based on their clinicopathological features.

Concentration of the Patients with ALK-Positive Lung Cancer

Clinicopathological features

Because ALK-positive lung cancer constitutes less than 5% of all lung cancers (Sasaki et al., ), it is critically important to select those patients who are more likely to have the ALK mutation. We initially attempted to select the patient population based on their clinicopathological features. As shown in Table 1, the patients with ALK-positive cancers were characteristically younger (a median age of 56 vs. 63), more frequently female, and more frequently non-smokers. In addition, the ALK fusion has a mutually exclusive relationship with the EGFR, KRAS, and HER2 mutations. Therefore, we attempted to select the ALK-positive lung cancer patients based on these clinicopathological features. The prevalence of the ALK translocation in the female patients younger than 63 without the EGFR, KRAS, HER2, or p53 mutations was 25% (Figure 1). However, four patients were not included in this subset (Mitsudomi et al., ). The selection of patients using clinicopathological features alone was able to increase the prevalence of the ALK mutation in our sample but was not able to identify all of the patients with it.

Figure 1

).

Morphological features

An alternative method of selection uses the morphological features of the tumors. Table 2 contains several morphological features known to be characteristic of ALK-positive adenocarcinomas (Figure 2). These characteristics have been identified in studies comparing ALK-positive and ALK-negative lung cancers (Inamura et al., ; Rodig et al., ; Takeuchi et al., ; Yoshida et al., ), mostly using surgical specimens. Among these features, the presence of signet ring cells has the highest predictive value for ALK-positive lung cancer. According to Rodig et al. () the ALK fusion was detected in 12 of the 26 tumors that they identified as having greater than 10% signet cells by area, although such tumors constituted only 8% of total lung adenocarcinomas. By contrast, the other morphological characteristics were more frequently seen in ALK-negative tumors. The solid-subtype was significantly more frequent in the ALK-positive cancers; however, the ALK-positive rate was 8% among the solid-subtype adenocarcinomas.

Figure 2

) found that 71% of the adenocarcinomas that consisted of more than 10% signet ring cells harbored the ALK fusion. One cellular characteristic of ALK-positive lung cancer is mucin production. In extreme cases, the alveolar space is filled with mucin (D), although the lumens of the neoplastic glands seldom open to the alveolar spaces. Although the majority of mucinous tumors are negative for TTF-1, the ALK-positive tumors (i.e., the signet ring cell carcinomas) are an exception.

Table 2

Growth pattern
   Acinar
   Cribriform
   Solid
Differentiation/grade
   Poorly differentiated
Cellular features
   Signet ring cell carcinoma
Cytoplasmic features
   Mucin production

Morphological characteristics of the ALK-positive tumors.

In practice, candidate patients for ALK inhibitor treatments typically have advanced cancer, and their biopsy specimens and/or cytology are the major source of samples. Therefore, predicting these patients’ ALK fusion status based solely on morphological features is notably difficult.

Three Key Methods of Detecting ALK Fusion

Currently, the methods of RT-PCR, FISH, and immunohistochemistry have been used to detect the ALK fusion gene. Because each method inevitably has both advantages and disadvantages (Table 3), we should be aware of their characteristics before applying them to clinical samples.

Table 3

ProsCons
RT-PCRA potentially rapid diagnostic methodDifficult to obtain high-quality of RNA
Very sensitiveNot applicable for unknown partners
More accurateDifficult to confirm the presence of tumor cells
Difficult to apply to archival tissues
FISHApplicable for any partnersExpensive
Screening method in clinical trialsRelative long turnaround time
Established in many labsLess sensitive
Applicable to archival tissues
IHCApplicable for any partnersIndirect demonstration of the fusion gene
Rapid turnaround timeOccasional false negative results
Established in many labsHigh dependence on antibody clones and detection methods
Applicable to archival tissues
Cheap

Advantages and disadvantages of ALK detection methods.

RT-PCR

RT-PCR is technically easy and rapid. The chromosomal inversion that characterizes the ALK fusion makes the sequence unique, and the PCR primer only hybridizes with the fusion chimeric transcript. This unique primer is responsible for the high sensitivity of this method. Soda et al. () detected fusion mRNA in sputum that contained as few as 10 EML4–ALK-expressing BA/F3 cells. This method is used only for known fusion partners, however, and all 11 of the reported variants require skillful application of the technique because of the long PCR products. In addition, high-quality RNA is difficult to obtain in clinical practice. We found that 4 of the 361 tumors we screened had inconsistent RT-PCR and immunohistochemical results. Three of the tumors revealed both new and known variants of the EML4–ALK fusion transcript with 5′-RACE. Therefore, this method is difficult to implement in a routine clinical diagnostic laboratory. However, RT-PCR is one of the few methods that provide direct evidence of the chromosomal translocation. A demonstration of chimeric transcription is the best direct evidence of the translocation; if the results are negative, however, it is clinically difficult to determine whether the tumor is truly negative for the ALK fusion or whether it was not detected due to impaired RNA integrity or technical errors.

Fluorescent in situ hybridization

Fluorescent in situ hybridization is currently the gold standard method used in clinical trials to detect the ALK fusion gene, and it was the first FDA-approved method for detecting the ALK fusion (Kwak et al., ). The FISH technique has been established in several labs to perform the HER2 test for breast cancer. FISH has the significant advantage of allowing archival material to be used. FISH also has disadvantages, however, including a relatively high cost and a long turnaround time. The evaluation of positive signals also requires considerable skill, especially when using biopsy samples. A break-apart FISH probe has been used to detect the ALK fusion (Figure 3), and the probes are designed for the telomeric and centromeric sides of the break points. Therefore, this design is applicable in detecting any kind of fusion partner, and any type of ALK gene rearrangement could theoretically be detected using this technique. However, it is known that some tumors with RT-PCR proven and ALK immunohistochemistry (IHC)-positive ALK fusions show non-split signals under the current criteria (Figure 4).

Figure 3

Figure 4

Immunohistochemistry

In contrast, immunohistochemical analysis is technically easy because it is integrated into routine pathological diagnosis. This assay has the advantages of rapidity, allowing the use of archival tissues, the ability to detect any partner genes, and simple comparisons of morphology. Although the immunohistochemical technique does not detect the ALK fusion gene itself, ALK is not detectable in any normal tissues other than the brain (Iwahara et al., ). Therefore, an ALK-positive reaction is associated with dysregulated expression of the gene, due to the altered promoter activity that is highly characteristic of an ALK inversion.

The early studies of ALK fusion immunohistochemistry reported that not all fusion-positive tumors yielded a positive immunohistochemistry result (Martelli et al., ). In contrast to the ALK fusion in anaplastic large cell lymphoma, the mRNA expression is lower in lung cancer; therefore, the false negative results appear to be caused by lower sensitivity. In general, the detection threshold is determined by the affinity of the primary antibody and the signal amplification system. In terms of the antibody affinity, Mino-Kenudson et al. () compared two clones of the anti-ALK antibody and found large differences in their affinities for lung cancer cells. Nevertheless, it has been reported that highly sensitive immunohistochemistry is quite effective at detecting the ALK fusion gene. Takeuchi et al. () compared the immunohistochemical results of combining three anti-ALK antibodies and two detection systems and found that all of the ALK-positive tumors could be detected by their highly sensitive immunohistochemical system, regardless of the differences between the clones. These two key factors, the antibody clones and the detection system, are important for detecting the ALK fusion by immunohistochemistry (Mitsudomi et al., ). Indeed, when we compared the results obtained using the ALK1 antibody with a conventional detection system to those obtained using the (high-affinity) 5A4 antibody with a highly sensitive detection system (Envision FLEX+ system), we found that 3 of 12 ALK-positive tumors went undetected by the former method (Table 4; Figure 5). 5A4 and D5F3 are known to be high-affinity antibody clones. When a highly sensitive detection system was used, either clone was able to detect all of the ALK-positive cancers, although D5F3 caused membranous staining in some ALK-negative tumors (Table 5).

Figure 5

Table 4

5A4 positive5A4 negative
TISSUE MICROARRAY (n = 361)
ALK1 positive100
ALK1 negative3348
BIOPSY (n = 43)
ALK1 positive30
ALK1 negative337

Comparison between conventional (ALK1 and standard ABC methods) and optimized IHCs (5A4 and Flex+ system) using tissue microarray and biopsy specimens.

Table 5

D5F3
5A4PositiveNegativeMembranous only
Positive1210943
Negative12343264

Comparison of IHC results according to the difference of antibody clones (5A4 and D5F3).

1Chimeric transcripts of EML4–ALK were detected in all cases.

2The tumor was negative for chimeric transcript of EML4–ALK and FISH.

3The tumors were negative for chimeric transcript of EML4–ALK.

Because IHC does not directly demonstrate ALK fusion, there are certain pitfalls. Certain small cell lung cancers have positive reactions (Figure 6) but do not have the ALK translocation. This positive reaction did not appear to be associated with gene amplification, and the reason for the positive reaction was unknown. The ALK protein may be expressed in association with the neuroectodermal differentiation of small cell lung cancer, as it is expressed in the normal brain (Iwahara et al., ). Another pitfall is a certain rate of false negative reactions in signet ring cells, even when using a high-affinity antibody with a highly sensitive detection method. A large amount of cytoplasmic mucin can often push the cytoplasm into the rim, and the thinned cytoplasm weakly demonstrates a positive reaction in certain cases (Figure 7). Fortunately, pure signet ring cell carcinoma is extremely uncommon in lung cancer, and the other tumor components can show positive reactions, or a FISH analysis can be used instead.

Figure 6

Figure 7

ALK Testing Guidelines by the Japanese Lung Cancer Society

The ALK inhibitor crizotinib is expected to be introduced in Japan, and we face the practical application of the methods described above because this agent will be approved for patients with ALK fusion detected using any methods. This is in contrast with US policy, where crizotinib has been approved for ALK-positive patients only with an FDA-approved test. As noted, different precautions are necessary for individual methods as well as the appropriate handling of the tissues. Under the circumstances, the Japanese Lung Cancer Society issued a guideline for ALK testing (Figure 8). The guidance mentioned detailed the pros and cons of these methods, precautions for tissue handling, tissue application (biopsy, cytology, and surgical specimens) and comparisons of the three methods, including the different results of the various antibody clones against ALK for immunohistochemistry. Although validation of this algorithm has recently begun in a large prospective cohort, it is stressed that concurrent multiple methods should be used to select ALK-positive patients.

Figure 8

Conclusion

The lung cancer with ALK fusion constitutes only a few percentages in overall lung cancer, thus the target proportion is very limited in comparison with the other molecular targeted treatments. However, if we succeed to achieve successful screening of the patients, the treatment with ALK inhibitor will be the most representative personalized therapy in terms of selecting the patients from those harboring the same category of tumors according to the infrequent gene alteration. In other words, the achievement may be a new challenge for the current clinical oncology.

Statements

Acknowledgments

The authors thank Noriko Shibata, Motoko Nimura, and Naoko Satoh for their excellent technical assistance with molecular, genetic, and immunohistochemical analyses.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

References

Summary

Keywords

anaplastic large cell lymphoma kinase, non-small cell lung cancer, immunohistochemistry, RT-PCR, FISH, screening, molecular targeted drugs

Citation

Murakami Y, Mitsudomi T and Yatabe Y (2012) A Screening Method for the ALK Fusion Gene in NSCLC. Front. Oncol. 2:24. doi: 10.3389/fonc.2012.00024

Received

05 December 2011

Accepted

19 February 2012

Published

16 March 2012

Volume

2 - 2012

Edited by

Giuseppe Giaccone, National Institutes of Health, USA

Reviewed by

Xu-Dong Zhu, McMaster University, Canada; Junfang Ji, National Cancer Institute, USA

Copyright

*Correspondence: Yasushi Yatabe, Department of Pathology and Molecular Diagnostics, Aichi Cancer Center, 1-1 Kanokoden, Chikusa-ku, Nagoya, Japan. e-mail:

This article was submitted to Frontiers in Cancer Molecular Targets and Therapeutics, a specialty of Frontiers in Oncology.

Disclaimer

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.

Outline

Figures

Cite article

Copy to clipboard


Export citation file


Share article

Article metrics