OUP user menu

Adenosquamous Carcinoma of the Lung
A Microdissection Study of KRAS and EGFR Mutational and Amplification Status in a Western Patient Population

Naobumi Tochigi MD, PhD, Sanja Dacic MD, PhD, Marina Nikiforova MD, Kathleen M. Cieply CISp(CG), Samuel A. Yousem MD
DOI: http://dx.doi.org/10.1309/AJCP08IQZAOGYLFL 783-789 First published online: 1 May 2011


Molecular testing of pulmonary adenocarcinomas for EGFR and KRAS mutations is becoming more common as tyrosine kinase inhibitor therapy is used for EGFR-mutated adenocarcinomas. Adenosquamous carcinomas represent a hybrid carcinoma, and there is no literature addressing the frequency of EGFR and KRAS mutations in this subset of lung carcinomas in Western populations. For this study, 23 adenosquamous carcinomas were microdissected with the glandular and squamous components analyzed for EGFR and KRAS mutations and EGFR amplification. In 3 cases (13%), there were EGFR mutations, with 2 having the identical mutation in the glandular and squamous elements. In 3 cases (13%), there were KRAS mutations in both histologic elements. Great heterogeneity existed in the rates of EGFR amplification in the 2 histologic components. Amplification was most common in both glandular and squamous components (11/23 [48%]). EGFR mutations occur in adenosquamous carcinoma in the same percentages as in conventional adenocarcinoma in the Western population. KRAS mutations are less common.

Key Words:
  • Adenosquamous carcinoma
  • EGFR
  • KRAS
  • Mutation
  • Amplification

Adenosquamous carcinoma of the lung is a rare subtype of non–small cell carcinoma of the lung, constituting 0.4% to 4% of cases. The definition of adenosquamous carcinoma indicates a carcinoma showing components of adenocarcinoma and squamous cell carcinoma, with each comprising at least 10% of the tumor.1 The prognosis of adenosquamous carcinoma is generally worse than that of adenocarcinoma and squamous cell carcinoma of the lung; however, new therapeutic options have been found for adenocarcinoma of the lung that could potentially impact adenosquamous lung carcinoma.2,3

The development of EGFR-targeted therapies, including monoclonal antibody (cetuximab) and tyrosine kinase inhibitors (gefitinib and erlotinib) have prolonged progression-free survival in high-stage adenocarcinomas of the lung.4 Responders to EGFR tyrosine kinase inhibitors often have somatic mutations in the EGFR tyrosine kinase domain, the most common being in frame deletions in exon 19, followed by a point mutation (CTG to CGG) in exon 21 at nucleotide 2573 that results in substitution of leucine by arginine at codon 858 (L858R).57

No effect has been consistently observed in squamous cell carcinomas of the lung in which EGFR mutations are absent.8 Clinical studies have also shown that EGFR fluorescence in situ hybridization (FISH)+ cases, compared with EGFR FISH– cases, are associated with favorable clinical benefits, including clinical response, stable disease, time to progression, and survival in patients with advanced adenocarcinoma treated with EGFR tyrosine kinase inhibitors.912 In contrast, KRAS mutations in pulmonary adenocarcinomas are resistant to EGFR tyrosine kinase inhibitor therapy.13 These mutations lead to substitutions of amino acids for glycine at positions 12 and 13 and can be identified in up to 30% of pulmonary adenocarcinomas. EGFR and KRAS mutations are mutually exclusive.7,1416

Because of its biphasic composition, adenosquamous carcinoma poses a therapeutic and molecular puzzle: Does it show a molecular profile more akin to adenocarcinoma or squamous cell carcinoma with regard to EGFR mutation and amplification and KRAS mutation, and how might this affect therapeutic decision making? All previous investigations have focused on Asian populations, which have different demographics and genetics than do Western populations. In this study, we reviewed the world literature on EGFR and KRAS alterations in primary adenosquamous carcinoma of lung in Asian populations and compiled the first analysis of 23 microdissected cases of well- to moderately differentiated adenosquamous carcinoma in a Western population, focused specifically on these genetic alterations and their implications for pathogenesis and therapeutic intervention.

Materials and Methods

Case Selection

We reviewed the Department of Pathology files at the University of Pittsburgh Medical Center, Pittsburgh, PA, from 1997 to 2009. We identified 23 cases of well- to moderately differentiated adenosquamous carcinoma, as defined by the 2004 World Health Organization classification of lung carcinomas.1 Poorly differentiated adenosquamous carcinomas were excluded because the distinction of glandular and squamous elements was too difficult to allow confident microdissection. Clinical information, including sex, age, smoking history, pathologic stage, and surgical procedure were obtained from the review of electronic medical records, with careful exclusion of all first- and second-generation Asian immigrants. Specimens included 4 wedge resections, 2 segmentectomies, 16 lobectomies, and 1 pneumonectomy. Two pathologists (N.T. and S.A.Y.) reviewed H&E-stained histologic sections and selected unequivocal glandular and squamous components for manual microdissection.

EGFR and KRAS Mutational Analysis

Tumor targets were manually microdissected from the 4-μm unstained histologic sections. DNA was isolated from each target using the DNeasy tissue kit (Qiagen, Valencia, CA) according to the manufacturer’s instructions. For the detection of mutations, DNA was amplified with primers flanking exon 19 of the EGFR gene (forward primer, 5′-CCCAGCAATATCAGCCTTAGGTG-3′ and reverse primer, 5′-CCACTAGAGCTAGAAAGGGAAAGAC-3′), exon 21 of the EGFR gene (forward primer, 5′-CCTCACAGCAGGGTCTTCTC-3′ and reverse primer 5′-CCTGGTGTCAGGAAAATGCT-3′), and exon 2 of the KRAS gene (forward primer, 5′-GGTGAGTTTGTATTAAAAGGTACTGG-3′ and reverse primer 5′-TCCTGCACCAGTAATATGCA-3′). Then, PCR products were sequenced in the sense and antisense directions using the BigDye Terminator version 3.1 cycle sequencing kit on the ABI 3130 (Applied Biosystems, Foster City, CA) according to the manufacturer’s instructions. The sequences were analyzed using Mutation Surveyor software (SoftGenetics, State College, PA). Each case was classified as positive or negative for the EGFR and KRAS mutations based on the sequencing results.


EGFR FISH analysis was carried out using the standard method with the dual-color EGFR SpectrumOrange/CEP7 SpectrumGreen probe (Abbott Molecular, Des Plaines, IL).4 In brief, paraffin sections were deparaffinized, dehydrated in ethanol, and air dried. The sections were digested with pepsin (0.5 mg/mL) at 37°C for 28 minutes. The tissue was denatured at 75°C for 5 minutes and dehydrated in ethanol. The probes were denatured for 5 minutes at 75°C before hybridization. Slides were hybridized overnight at 37°C and washed in 2× saline sodium citrate/0.3% NP40 at 72°C for 2 minutes. The nuclei were counterstained with DAPI I/antifade (Abbott Molecular).

Each FISH assay included normal lung tissue sections as negative control samples and sections of non–small cell lung carcinoma previously identified as EGFR FISH+ as positive control samples. Analyses were carried out using a fluorescence microscope (Nikon Optiphot-2 and Quips Genetic Workstation) equipped with a Chroma Technology 83000 filter set (Chroma Technology, Bellows Falls, VT) with single-band exciters for Texas red/rhodamine, fluorescein isothiocyanate, and DAPI (UV 360 nm). The histologic areas previously selected on the H&E-stained sections were identified on the FISH-treated slides. Only individual and well-delineated cells were scored. Overlapping cells were excluded from the analysis. At least 60 cells were scored for each case and control.

EGFR FISH results were interpreted by using the Colorado scoring system, as previously described.12 In brief, tumors with EGFR gene amplification defined as an EGFR gene/CEP7 ratio of 2 or more or with at least 40% of cells showing at least 4 copies of the EGFR signals were classified as EGFR FISH+. Tumors with fewer than 40% of cells showing at least 4 copies of the EGFR signals and an EGFR gene/CEP7 ratio of less than 2 were classified as EGFR FISH–.12


Clinical and pathologic data are shown in Table 1. Of the patients, 16 were men and 7 were women. The average age at diagnosis was 71 years (range, 53–82 years). The average diameter of the tumor was 2.8 cm (range, 1.0–9.0 cm). Mutational analysis showed 3 cases (13%) with a KRAS gene mutation, 3 cases (13%) with an EGFR mutation, and 17 cases (74%) negative for KRAS and EGFR mutations Image 1. KRAS mutations in codon 13 (p.G13D) were found in the glandular and squamous components in 2 cases, and a mutation in codon 12 (p.G12C) was also found in both components in 1 case. However, the mutant peak in the glandular component was 100% higher than the normal peak in the squamous component, reflecting a difference in mutation dosage in this third case. Among the cases with an EGFR mutation, all cases occurred in women and had exon 19 deletions: 1 case had an in frame 15-base-pair deletion (p.E746_A750del) of both components, 1 case had a novel point mutation at amino acid 747 leading to substitution of leucine to proline (p.L747P) of both components, and 1 case had a homozygous polymorphism (p.I744I) without an amino acid substitution in the glandular element Image 2. No tumor had mutations of EGFR and KRAS. In summary, in two thirds of the cases with EGFR and KRAS mutations, the mutations were the same in the glandular and squamous components; in one third, they differed.

The relationship between EGFR/KRAS gene mutation and EGFR FISH positivity is shown in Table 2. By using the Colorado scoring criteria, we classified 11 cases (48%) as positive and 5 cases (22%) as negative in both components. EGFR mutations were associated with EGFR amplification, but specificity was limited because amplifications were also seen in adenosquamous carcinomas without EGFR mutations and even in cases with KRAS mutations.

View this table:
Table 1
Image 1

Examples of adenosquamous carcinoma with KRAS mutations. In the glandular element (A, H&E, ×200; B; and C), a normal KRAS point mutation was identified (B). EGFR amplification was negative (C). In the squamous component (D, H&E, ×200; E; and F), there was a difference in the dosage of mutation (E). EGFR fluorescence in situ hybridization positivity was defined in part by larger and brighter EGFR signals (red) than CEP7 signals (green) (F).

Image 2

Examples of adenosquamous carcinoma with EGFR mutations. In the glandular element (A, H&E, ×200; B; and C), an EGFR exon 19 homozygous polymorphism was identified (B). EGFR amplification was negative (C). In the squamous element (D, H&E, ×200; E; and F), the EGFR mutation was negative (E). EGFR fluorescence in situ hybridization positivity was defined by EGFR gene amplification in this case (F).


Adenosquamous carcinoma of the lung is a rare primary lung carcinoma, with few studies analyzing its molecular profile.3 Almost all studies published to date have occurred in Asian populations, and no study has analyzed a Western population with adenosquamous carcinoma Table 3.1725 This study represents the first study to assess adenosquamous carcinomas by microdissecting the glandular and squamous components and analyzing them separately for KRAS and EGFR mutations and EGFR copy number changes by FISH. Several features are worth noting.

View this table:
Table 2

In contrast with studies in Japanese patients, we did not find a significant difference in the frequency of EGFR mutations between adenosquamous carcinomas (13%) and typical adenocarcinomas in the Western population (10%).26 Sasaki et al23 reported only 4 (15%) of 26 adenosquamous carcinomas positive for EGFR mutations and, in the same period, a much higher frequency of 40.3% (146/362) in patients with adenocarcinoma. However, Kang et al,18 in a Korean population, found similar frequencies of EGFR mutations in adenosquamous carcinomas (44%) and adenocarcinomas. This striking discrepancy between these 2 Asian studies may reflect a difference in methods of mutational analysis. Kang et al18 used a more sensitive polymerase chain reaction (PCR) single-strand conformation polymorphism (SSCP) method, whereas Sasaki et al,23 similar to our study, used less sensitive direct DNA sequencing. However, the possibility of genetic differences between Korean and Japanese populations cannot be entirely excluded.

In all Asian studies, as well as in our study, EGFR mutations occurred only in women. Similar to the study by Sasaki et al,23 all EGFR-mutated cases in our study had exon 19 mutations with no exon 21 mutations. Only 1 mutation in exon 19 was a typical deletion found in adenocarcinomas. The 2 other mutations were a novel point mutation and a homozygous polymorphism. Although our patient population is certainly small, these findings are somewhat different from the expected findings in most adenocarcinoma series in which exon 21 mutations are represented. In contrast, Kang et al18 by using PCR-SSCP were able to identify 1 case with an exon 21 mutation. This observation again brings up the problem of sensitivity of direct DNA sequencing in the detection of EGFR mutations, particularly point mutations, in comparison with more sensitive methods such as PCR-SSCP. Nevertheless, our observations simply may be a consequence of the small number of cases in our study, and it would be worthwhile studying a larger case series. A high frequency of exon 19 deletions is intriguing because patients with adenocarcinoma with exon 19 deletions had worse survival than patients with exon 21 mutations.27

View this table:
Table 3

Adenosquamous carcinomas are associated with a worse prognosis regardless of the patient’s ethnic background. Our study does not have follow-up data because that was not the aim, but it would be interesting to further study the prognostic importance of exon 19 deletions in adenosquamous carcinomas.

It is interesting that KRAS mutations, seen in approximately 30% of Western lung adenocarcinomas, were identified in only 13% of our adenosquamous cases, suggesting that KRAS mutations may not have as significant a role in adenosquamous carcinoma as in primary adenocarcinomas of the lung occurring in smokers.26 These results need to be interpreted with some caution because our numbers are small, and the vast majority of our cases involved smokers and may not be directly comparable to “typical” adenocarcinoma cases. It should be noted, too, that their incidence is still greater than that expected in squamous carcinoma of the lung. KRAS mutations were uncommon events in Asian populations as well.18,23 The results of EGFR and KRAS mutational analysis in Asian and Western populations suggest that carcinogenesis of adenosquamous carcinomas, in contrast with adenocarcinomas, may be more similar between these 2 ethnic groups. However, larger studies using the same mutational assay would be necessary to confirm this assumption.

The molecular assessment of adenosquamous carcinoma also raises the issue of clonal evolution of this non–small cell carcinoma. Historically, investigators have speculated as to whether mixed carcinomas arise from a common progenitor cell or whether such carcinomas represent collision tumors derived from separate progenitors. Although this study was not designed to study this question, it is clear that when looking at EGFR and KRAS mutations, the glandular element and squamous component of adenosquamous carcinoma had these same EGFR and/or KRAS mutations in 67% of our cases. This would argue for a common progenitor cell, as suggested in some prior studies.3,18 It highlights that when one observes glandular differentiation in a tumor with apparent squamous cell differentiation, it would be worthwhile to assess either element of such tumors for molecular alterations because this subset of non–small cell carcinoma can have EGFR mutations (unlike classical squamous carcinoma) and could be amenable to tyrosine kinase inhibitor therapy.

The interpretation of EGFR FISH results in adenosquamous carcinoma is difficult because of differences in the criteria for amplification and because of the biphasic cell populations. Clearly the literature indicates that EGFR amplification is more frequent in squamous cell carcinoma than in adenocarcinoma.28 In our study, the squamous component demonstrated more EGFR amplification than the glandular elements on a case-by-case basis, reflecting other pathways for EGFR amplification than EGFR mutation. Amplifications using the Colorado criteria were sensitive but not specific for EGFR mutations. It is interesting that EGFR amplification using the Colorado criteria was also seen in cases without EGFR mutation but with KRAS mutation, a previously reported finding.29 These findings provide some support for our belief and that of others that mutations of EGFR tend to correlate with amplification and to better predict response to interventional therapy. Similar to adenocarcinomas of lung, EGFR amplification most likely represents a late event in carcinogenesis of adenosquamous carcinoma, regardless of the initiating mutational events.30

There is scant literature on mutational studies of adenosquamous carcinomas outside the lungs.31,32 In uterine cervical adenosquamous carcinoma, no activating mutations in the hotspot region (exons 18–21) of the EGFR gene were identified. It is interesting that a silent base substitution in EGFR exon 20 was found in 56% of cases.32 In pancreatic adenosquamous carcinoma, heterozygous KRAS mutations were found in all 8 cases. Like our studies, some variability was noted in the squamous and glandular elements.31

There are notable limitations to this study. First, it is a retrospective study with a relatively small number of cases, although adenosquamous carcinoma of the lung is rare. Second, we restricted this study to well- to moderately differentiated tumors and did not include poorly differentiated tumors that could be controversial diagnostically and are very difficult to microdissect to obtain isolated cell populations. Although we saw no obvious molecular differences in differentiated tumors, poorly differentiated tumors could vary. Third, because of the recent discovery of tyrosine kinase inhibitor therapy, we do not have data on the potential response of these tumors; this will need to be accomplished in prospective studies.

In this microdissection study of adenosquamous cell carcinoma, we identified heterogeneity not only in cellular morphologic features, but also in mutational analysis and EGFR gene copy numbers. We noted that adenosquamous cell carcinoma had a similar EGFR mutational rate to adenocarcinoma of lung but interestingly showed a KRAS mutation rate approximately half of that expected in our Western patient population. When EGFR and KRAS mutations were seen in the glandular element, such mutations were also observed in the squamous component in two thirds of cases. In contrast, EGFR amplification using the Colorado criteria was more variable and less predictive of mutation of KRAS and EGFR mutations. In this rare form of lung carcinoma, we believe that molecular assessment predominantly by mutational analysis may be warranted as investigators seek alternative therapies.


We thank Diana Winters for secretarial and administrative assistance and Linda Shab and Tom Bauer for photographic assistance.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
View Abstract