Presented in part at the 59th Annual Meeting of American Society of Cytopathology; November 4-8, 2011; Baltimore, MD.
The identification of molecular alterations has an important therapeutic implication in patients with lung adenocarcinomas. In the current study, the authors evaluated their experience with the identification of epidermal growth factor receptor (EGFR), Kirsten rat sarcoma viral oncogene homolog (KRAS) mutation, and anaplastic lymphoma kinase (ALK) gene rearrangement using cytological specimens of primary and metastatic lung adenocarcinoma.
A total of 54 cases of lung adenocarcinomas (11 primary and 43 metastatic tumors) in which molecular tests were performed were retrieved. Molecular tests were performed on the cell block material of 19 effusions and 35 fine-needle aspirates. EGFR mutation was evaluated by polymerase chain reaction sequencing analysis of exons 18, 19, 20, and 21. KRAS mutation was tested using polymerase chain reaction–single-strand conformational polymorphism analysis of codons 12 and 13. ALK gene rearrangement was evaluated by fluorescence in situ hybridization using an ALK break apart probe.
Molecular tests were successful in 49 of 54 cases (91%). Evaluation of EGFR mutation, KRAS mutation, and ALK gene rearrangement were performed in 49 cases, 14 cases, and 22 cases, respectively. EGFR mutations were found in 14 of 49 cases (29%), including 5 primary and 9 metastatic tumors. Three metastatic/recurrent adenocarcinomas demonstrated an additional EGFR T790M mutation that was not identified in the original specimens. KRAS mutation was detected in 3 of 14 cases (21%) including 1 primary and 2 metastatic tumors. ALK gene rearrangement was evident in 3 of 22 cases (14%), all of which were metastatic tumors.
Lung cancer is the leading cause of cancer death both worldwide and in the United States. Non-small cell lung cancers (NSCLCs), primarily adenocarcinoma and squamous cell carcinoma, comprise approximately 85% of all lung cancers. Despite advances in early detection, the majority of patients with NSCLCs present at an advanced stage with unresectable disease. Among the remaining 15% to 20% of patients with resectable disease, many are poor surgical candidates due to significant comorbidities. In these nonsurgical candidates, cytological materials are often the only specimens available for diagnosis and further workup for tailoring systemic therapies. In light of the recently discovered molecular alterations in lung adenocarcinomas and the advent of targeted therapies, the need for further classification of NSCLCs and molecular testing on tissue gleaned from cytologic specimens cannot be emphasized enough.[1-4]
To our knowledge, several mutations have been identified to date in adenocarcinomas of the lung, most notably including the epidermal growth factor receptor (EGFR) mutation, Kirsten rat sarcoma viral oncogene homolog (KRAS) mutation, and echinoderm microtubule-associated protein-like 4 and anaplastic lymphoma kinase (EML4-ALK) translocation.[1, 2] Together, these molecular alterations are identified in slightly less than one-half of cases of lung adenocarcinoma. EGFR, a transmembrane growth factor receptor with a cytoplasmic tyrosine kinase domain, is perhaps the best-studied molecular target, resulting in the development of tyrosine kinase inhibitors such as gefitinib and erlotinib.[5-7] The KRAS gene codes for a signaling protein, which is downstream of the EGFR signal transduction pathway. Previous studies have suggested that the KRAS mutation is predictive of resistance to tyrosine kinase inhibitor therapy and a poorer prognosis. ALK gene rearrangement is a relatively less common translocation detected in lung adenocarcinomas. Molecular testing for the ALK gene rearrangement is becoming increasingly relevant because crizotinib, a small-molecule ALK inhibitor, has shown promising results in clinical trials for the treatment of lung adenocarcinomas harboring the ALK gene rearrangement.[9, 10] Therefore, the identification of these molecular alterations in lung adenocarcinomas would have profound therapeutic implications.
It has been recommended that molecular tests for NSCLCs be performed on biopsy tissues rather than cytological specimens.[11, 12] The issues associated with cytological specimens reside in the amount of material, the interference from background nonneoplastic cells, and the standardization of cytological preparations. Recent studies have documented the use of cytological specimens for identifying molecular alterations such as EGFR and KRAS mutations in NSCLCs with an equivalent or even higher sensitivity.[14-17] However, there are rare reports regarding the ALK gene rearrangement performed in cytology specimens. In the current retrospective study, we evaluated our experience with the identification of EGFR mutations, KRAS mutations, and ALK gene rearrangements using cytological materials from both primary and metastatic adenocarcinomas of the lung.
MATERIALS AND METHODS
We searched the cytopathology archives and identified 54 patients who had received a cytological diagnosis of lung adenocarcinoma and had at least 1 of the molecular tests performed at the study institution during the period from July 2008 to June 2011. These included 10 consultation cases (19%), which were sent to the study institution primarily for molecular testing. The patients were 25 men and 29 women ranging in age from 24 years to 90 years (mean age, 68 years). The cases included 11 primary lung adenocarcinomas (20%) and 43 metastatic lung adenocarcinomas (80%). The metastatic sites were pleural (17 cases), lymph nodal (16 cases), soft tissue (3 cases), pericardial (2 cases), bone (2 cases), adrenal (2 cases), and liver (1 case).
The specimens included 35 fine-needle aspirates and 19 effusion specimens. Direct smears, stained using both the Diff-Quik and Papanicolaou methods, were prepared from fine-needle aspiration specimens whereas Papanicolaou-stained, ThinPrep slides were prepared from effusion specimens for cytological evaluation. The remainder of fine-needle aspirates or percentage of effusion specimens were fixed in CytoRich Red fixative (BD Diagnostics, Franklin Lakes, NJ) and subjected to centrifugation at 800 revolutions per minute for 5 minutes. The resulting pellet was processed for a cell block using the HistoGel method (Thermo Fisher Scientific, Kalamazoo, Mich). The cell block was subjected to routine tissue processing including formalin fixation followed by paraffin embedding, the sections of which were stained with hematoxylin and eosin (H&E) for morphologic evaluation.
Specimen Selection for Molecular Testing
In the current study, all cases had a cytological diagnosis of lung adenocarcinoma, either primary or metastatic, that was rendered based on the cytomorphologic features and/or immunocytochemical studies. For primary lung adenocarcinomas, thyroid transcription factor-1 (TTF-1) and cytokeratin 5/6 (CK5/6) or p63 immunostains were performed in 6 of 11 cases (55%), primarily to differentiate adenocarcinoma from squamous cell carcinoma. TTF-1 and/or CK7/CK20 immunostains were performed to elucidate the tumor origin in 37 of 43 cases (86%) of metastatic lung adenocarcinomas. Twenty-six of 54 cases (48%) had concurrent or follow-up biopsies or excisions to confirm the cytologic diagnosis including 7 of 11 primary lung adenocarcinomas (64%). The remaining 4 primary lung adenocarcinoma cases had cytomorphologic features of adenocarcinoma and were found to be positive for TTF-1, thereby confirming a lung primary tumor.
Because all molecular tests were performed on the cell block sections, H&E-stained cell block sections were evaluated for tumor cellularity and tumor cell/nonneoplastic cell ratios. In general, the cases with > 100 tumor cells and a tumor cell/nonneoplastic cell ratio of > 40% underwent EGFR mutation testing. For those cases with a tumor cell/nonneoplastic cell ratio of < 25%, microdissection might be needed to enrich the tumor cells. It was estimated that < 10% of cases required microdissection, the majority of which were exfoliative specimens. For KRAS mutation and ALK gene rearrangement tests, a ratio of tumor cells/nonneoplastic cells as low as 5% was acceptable.
At the study institution, the oncologists generally request molecular tests in consultation with molecular pathologists. There is not a specific algorithm to follow but in general EGFR mutation analysis is performed. Based on the EGFR mutation status and clinicopathologic characteristics, additional molecular tests, KRAS mutation analysis, and/or ALK gene rearrangement analysis may or may not be performed. At the cytology service, a cytopathologist refers cases that have a diagnosis of non-small cell carcinoma/adenocarcinoma and shows the tumor cells on the cell block to a molecular pathologist. No blank sections are cut upfront. The molecular pathologist decides whether the cell block will be used for molecular testing based on the tests requested by oncologists and the quality and quantity of tumor cells on the cell block, as well as the availability of concurrent surgical biopsy specimens.
EGFR mutation analysis was performed by polymerase chain reaction (PCR) followed by Sanger sequencing using the procedure previously described. Briefly, DNA was extracted from cell block sections using the Qiagen kit (Qiagen, Valencia, Calif). Exons 18 to 21 of EGFR were amplified by nested PCR, followed by direct Sanger sequencing. Sanger sequencing was performed in all EGFR cases, with the exception of exon 19, which was analyzed initially by running the PCR product with size comparison with that of the wild-type. When short production was detected (possible deletional mutation), the PCR product was then Sanger sequenced to confirm the mutational status.
KRAS mutation analysis was performed by PCR–single-strand conformational polymorphism (PCR-SSCP) using the procedure previously described.[18, 19] The PCR-SSCP has been shown to have an analytical sensitivity of 5% for identifying KRAS mutations. Briefly, DNA was extracted from cell block sections using the Qiagen kit. The codons 12 and 13 of exon 2 were subjected to conventional PCR amplification. The resulting PCR product was separated by electrophoresis on nondenaturing gel. The presence of the KRAS mutation was determined by comparing the SSCP patterns with those positive controls that have known mutations. In cases with an unknown mutation based on the gel pattern, direct sequencing of corresponding codons was performed.
ALK gene rearrangement was accomplished by fluorescence in situ hybridization using an ALK break apart probe (Abbott Laboratories, Abbott Park, Ill) following the procedure previously described. Fluorescent labels were affixed to the 250-kilobase pair and 300-kilobase pair regions flanking the ALK gene breakpoint region. A positive ALK gene rearrangement result was defined as a separation of red and green probes whereas a negative translocation was defined as a yellow signal (produced by overlapping red and green signals). A positive signal observed in > 15% of the tumor cells was used as the cutoff for a positive ALK gene rearrangement. At least 50 tumor cells with informative hybridization signals were required for an appropriate interpretation.
Data were analyzed as the percentage of positive results for each test and the results were compared between primary and metastatic tumors.
EGFR mutation, KRAS mutation, and ALK gene rearrangement molecular analyses were performed in 54 cases, 14 cases, and 22 cases, respectively. Of the 54 cases that underwent EGFR mutation analysis, 5 cases (9%) were inadequate for analysis due to insufficient materials. These inadequate cases included 3 fine-needle aspiration and 2 pleural effusion specimens.
Of 49 cases with a complete EGFR mutation analysis, a positive EGFR mutation was identified in 14 cases (29%) (Table 1) (Fig. 1). The detected EGFR mutations included an L858R point mutation in exon 21 (5 cases), a deletional mutation in exon 19 (6 cases), and a T790M point mutation in exon 20 (3 cases) (Table 2). Of the 14 cases tested for the KRAS mutation, 3 (21%) demonstrated positive results (Table 1) (Fig. 2). A TGT mutation in codon 12 was seen in 2 cases whereas 1 case showed an AGT mutation in codon 12. ALK gene rearrangement was identified in 3 of the 22 cases (14%) tested (Table 1) (Fig. 3). The molecular alterations identified in the current study (EGFR mutation, KRAS mutation, or ALK gene rearrangement) appeared to be mutually exclusive in all lung adenocarcinomas.
Table 1. EGFR Mutation, KRAS Mutation, and ALK Gene Rearrangement In Primary And Metastatic Lung Adenocarcinomas
The EGFR mutation appeared to occur with a higher frequency in primary lung adenocarcinomas compared with their metastatic counterparts (45% vs 24%) (Table 1). The distribution of specific EGFR mutations in primary and metastatic lung adenocarcinomas is detailed in Table 2. There was no significant difference in the frequency of KRAS mutation noted between primary and metastatic lung adenocarcinomas. Positive ALK gene rearrangement was identified in 3 cases, all of which were metastatic lung adenocarcinomas (Table 1).
There were 8 cases in which at least 1 of the molecular tests was tested in both the original primary and metastatic/recurrent tumors (Table 3). A positive EGFR deletional mutation in exon 19 or a point mutation in exon 21 was found in 5 primary tumors and also was identified in the following metastatic/recurrent specimens. An additional EGFR T790M point mutation was presented in 3 of these 5 cases. The remaining 3 cases indicated a negative EGFR mutation in both primary and metastatic/recurrent tumors. In addition, ALK gene rearrangement was found to be positive and negative in 1 case each of metastatic/recurrent tumors.
Table 3. Molecular Alterations in Matched Original Primary and Recurrent/Metastatic Tumors
The data from the current study demonstrated that EGFR mutation, KRAS mutation, and ALK gene rearrangement tests can be performed on the cytological specimens with an overall success rate of 91%. In cases combining primary and metastatic lung adenocarcinomas, EGFR mutations, KRAS mutations, and ALK gene rearrangement were identified in 14 of 49 cases (29%), 3 of 14 cases (21%), and 3 of 22 cases (14%), respectively. Furthermore, all the molecular alterations identified were mutually exclusive. These results are similar to those previously reported in the literature,[1, 2] but the positivity rates for EGFR mutations and ALK gene rearrangement are higher. Of the 14 positive EGFR mutation cases, a deletional mutation in exon 19, an L858R point mutation in exon 21, and a T790M point mutation in exon 20 were noted in 6 cases, 5 cases, and 3 cases, respectively. The reasons for the higher EGFR mutation rate noted in this study are uncertain. All cases included in the current study had a specific diagnosis of lung adenocarcinoma, not just NSCLC as reported in the majority of literature. This also may reflect the finding that cases in the current study were more selective because at the study institution a specific molecular test was often requested by clinicians in consultation with pathologists after reviewing the clinicopathologic characteristics of each individual case. This case serial included 10 consultation cases that were submitted primarily for molecular testing. Similarly, these clinicopathologic features may also help explain our higher rate of ALK gene rearrangement. In addition, the majority of cases in the current study were metastatic tumors. It has been suggested that lung adenocarcinomas harboring ALK gene rearrangements are associated with more aggressive clinical behavior. It should be no surprise that ALK gene rearrangement is more frequently observed in metastatic tumors. Indeed, in the current study, all 3 cases harboring an ALK gene rearrangement were metastatic tumors. However, these were just our observations and are still considered to be inconclusive due to the limited number of cases in the current study. To the best of our knowledge, the exact frequency of these molecular alterations in primary versus metastatic tumors has yet to be explicitly studied and determined in a larger case series.
Currently, it has been recommended that molecular tests be performed on surgical specimens.[11, 12] However, cytological specimens can be obtained via much less invasive procedures and are sometimes the only materials available for the tests. Furthermore, methanol-based/ethanol-based fixatives are routinely used in the preparations of cytological specimens, which may allow for better DNA quality and thus higher success rates for molecular tests. Molecular testing of EGFR and KRAS mutations performed in a variety of cytological preparations has demonstrated equivalent or higher sensitivity and accuracy than surgical specimens.[14, 15] The issues associated with cytological specimens may include a limited amount of material, interference from background nonneoplastic cells, and a lack of standardization of cytological preparations. Several studies have emphasized the importance of the role of cytopathologists in determining the adequacy of samples as well as accurate subtyping of NSCLC (adenocarcinoma vs squamous cell carcinoma) to maximize the cost-effectiveness of molecular testing.[3, 14, 22] For the majority of our cases of fine-needle aspiration biopsy, a cytopathologist was available for rapid on-site evaluation to ensure adequate sampling. We used cell block materials for the molecular tests, which allowed for the accurate estimation of tumor cell numbers as well as tumor cells/nonneoplastic cell ratios. Implementation of screening adequate tumor cellularity and the tumor/normal cell ratio might improve results by ensuring an adequate amount of tumoral DNA is available. A variety of different tumor cellularity cutoff values have been used in the literature, ranging from 25% to 70% tumor cells; some authors have used cytopathologists to screen slides for molecular testing, whereas others have used different forms of microdissection.[3, 14, 15, 23] Accurate molecular testing can be achieved using as few as 30 to 100 tumor cells. Smouse et al have determined that molecular alterations can be most reliably detected with > 50% tumor cells. With adequate percentages of tumor cells (70%), testing for KRAS and EGFR using allele-specific quantitative PCR was sensitive enough to detect all mutations detected in surgical resection specimens. Nevertheless, molecular testing can still fail, even with the implementation of these quality assurance efforts. In the current study, 5 cases were initially believed to be sufficiently cellular to attempt molecular testing but were subsequently found to be inadequate. At the study institution, molecular tests are currently performed only when requested by oncologists and no blank sections are cut upfront. The failure of molecular testing noted in 5 cases in the current study was most likely due to tissue exhaustion when the cell blocks were cut at deeper levels. Thus, implementing a protocol in which blank sections are cut in between H&E-stained levels would minimize the possibility of tissue exhaustion and further improve the analytic phase of molecular testing.
The molecular signatures associated with lung adenocarcinoma may evolve when the tumor progresses or metastasizes. Previous studies of metastatic lung adenocarcinomas have yielded mixed results, with some studies indicating concordant and the others discordant mutational status in primary versus metastatic tumors.[25-28] The successful use of archived specimens offers the advantage of comparison between prior tumors and recurrences/metastases, which occasionally acquire mutations that are different from those of the original tumor. It has been known that specific EGFR T790M mutations (which rarely occur de novo), are commonly observed in patients who develop resistance after anti-EGFR or other systemic therapies. In the current study, there were 8 cases in which molecular tests were performed in both original primary tumors and those that were recurrent/metastatic. EGFR mutation status (5 positive and 3 negative cases) was the same in all the matched pairs. In addition, a new T790M mutation was identified in 3 recurrent/metastatic tumors with mutant EGFR. All 3 patients with tumors harboring an EGFR T790M mutation had clinically failed anti-EGFR therapy. Because of potential therapeutic implications, it is critically important to have repeated molecular tests performed in patients with metastatic tumors or in those with tumors that failed therapies. As demonstrated by the current study and others,[14-17] cytological specimens acquired by minimally invasive methods can adequately be used for molecular analysis. Indeed, serial molecular testing on small biopsy or cytological specimens instead of serial excisional biopsies is an emerging paradigm in the management of patients receiving molecular targeted therapies.
In summary, molecular tests such as EGFR mutation, KRAS mutation, and ALK gene rearrangement analyses can be reliably performed on cell block materials of cytological specimens from both primary and metastatic lung adenocarcinomas. Repeating molecular testing in patients with recurrent/metastatic tumors may be beneficial for tailoring systemic therapies. A limited quantity of tumor volume remains an issue for small biopsy specimens, including cytological specimens.