OUP user menu

Micropapillary Lung Adenocarcinoma
EGFR, K-ras, and BRAF Mutational Profile

Rosane De Oliveira Duarte Achcar MD, Marina N. Nikiforova MD, Samuel A. Yousem MD
DOI: http://dx.doi.org/10.1309/AJCPBS85VJEOBPDO 694-700 First published online: 1 May 2009

Abstract

Micropapillary lung adenocarcinoma (MPA) has been reported as an aggressive variant of adenocarcinoma, frequently manifesting at high stage with a poor prognosis. We analyzed the clinical and molecular profile of 15 primary MPAs for K-ras, EGFR, and BRAF mutations and performed fluorescence in situ hybridization for EGFR amplification. In our study, 11 (73%) of 15 MPAs harbored mutually exclusive mutations: 5 (33%) K-ras, 3 (20%) EGFR, and 3 (20%) BRAF. Mutations in all 3 genes occurred in patients with a smoking history and tumors with mucinous differentiation and secondary lepidic, acinar, and solid growth, suggesting that in a Western population, cytomorphologic correlation with genetic mutations is more unpredictable than in Japanese cohorts. We conclude that K-ras, EGFR, and BRAF mutations are disproportionately seen in adenocarcinomas of lung with a dominant micropapillary growth pattern compared with conventional adenocarcinoma in our institutional experience.

Key Words:
  • Micropapillary adenocarcinoma
  • Papillary adenocarcinoma
  • Bronchioloalveolar adenocarcinoma

Lung adenocarcinomas are histologically heterogeneous, which led the World Health Organization to highlight that most adenocarcinomas had a mixed growth pattern, with characteristic patterns being solid, acinar, papillary, and le-pidic.1 Lung adenocarcinomas with papillary growth show 2 types of papillary architecture: true papillary structures with papillae containing a layered glandular epithelium surrounding a fibrovascular core and micropapillary growth in which the papillary tufts lack a central fibrovascular core and extensively shed within alveolar spaces.2,3 Micropapillary growth patterns have been associated with an aggressive clinical course compared with traditional papillary adenocarcinoma and bronchioloalveolar carcinoma.210 Micropapillary adeno-carcinoma (MPA) often manifests at a high stage in nonsmokers, with intralobar satellites, and frequently metastasizes to the contralateral lung, mediastinal lymph nodes, bone, and adrenal glands, with high mortality.412

Because MPA represents a unique form of lung adenocarcinoma, we analyzed 15 MPA cases for the common genetic mutations in lung adenocarcinoma to determine whether a distinct genetic profile was associated with this histopathologic growth pattern. To our knowledge, no study comparing K-ras, EGFR, and BRAF mutations in micropapillary adenocarcinoma has been reported.

Materials and Methods

The 1997–2008 pathology files of the University of Pittsburgh Medical Center, Pittsburgh, PA, were searched for lung adenocarcinomas showing micropapillary growth. The study was approved by the institutional review board of the University of Pittsburgh Medical Center. Histologic classification was according to the revised World Health Organization classification of lung malignancies of 2004, with micropapillary adenocarcinoma defined according to the criteria of Silver and Askin13 for papillary adenocarcinoma, in which greater than or equal to 75% of the adenocarcinoma manifested a micropapillary growth pattern with primary and secondary budding from papillary stalks.1

In this series, micropapillary growth was always intimately admixed with a lesser component of papillary growth. Because the two were nonseparable, we used dominant micro-papillary growth, in a background of simple papillae, as part of the histologic definition of MPA. Of the 37 cases identified, 15 adenocarcinomas conformed to these criteria. There was an average of 5 tumor sections (range, 2–11 sections) of each adenocarcinoma available for review. The adenocarcinomas were analyzed by patient age, sex, and smoking history and by tumor location and stage. Tumors were also evaluated for the following morphologic features: tumor size, grade, secondary architectural growth patterns, and predominant cell type (hobnail vs columnar); goblet, clear cell, and signet-ring cell differentiation; mucin production; psammoma bodies; host inflammatory response; presence of necrosis; and visceral-pleural and angiolymphatic invasion.

Fluorescence in situ hybridization analysis for EGFR gene amplification was performed with the dual-color EGFR SpectrumOrange/CEP7 SpectrumGreen probe and paraffin pretreatment kit (Vysis, Downers Grove, IL) using previously described methods.14 Amplification was determined by the ratio of the number of EGFR signals per cell to the number of chromosome 7 centromere signals per cell. Amplification was defined as a ratio of 2.0 or greater.

Direct DNA sequencing of codons 12 and 13 of exon 2 of the K-ras gene and exons 19 and 21 of the EGFR gene was performed as previously described and according to the manufacturer’s instructions using the BigDye Terminator v3.1 cycle sequencing kit on an ABI 3130 (Applied Biosystems, Foster City, CA).1416 All polymerase chain reaction (PCR) products were sequenced in sense and antisense directions. The sequences were analyzed by using Mutation Surveyor software (SoftGenetics, State College, PA). Each case was classified as positive or negative for the K-ras and EGFR mutations based on the sequencing results.

Detection of the BRAF V600E mutation was performed using real-time PCR and post-PCR fluorescence melting curve analysis on a LightCycler (Roche Applied Science, Indianapolis, IN), as previously reported.16 Briefly, a pair of oligonucleotide primers flanking the mutation site was designed, together with 2 fluorescent probes, with the sensor probe spanning the nucleotide position 1799. Amplification was performed in a glass capillary using 50 ng of DNA in a 20-μL volume. The reaction mixture was subjected to 40 cycles of rapid PCR consisting of denaturation at 94°C for 1 second, annealing at 55°C for 20 seconds, and extension at 72°C for 10 seconds. Postamplification fluorescence melting curve analysis was performed by gradual heating of samples at a rate of 0.2°C/s from 45°C to 95°C. All PCR products that showed deviation from the wild-type (placental DNA) melting peak were sequenced to verify the presence of mutation.

Our internal study of more than 350 primary adenocarcinomas (exclusive of micropapillary adenocarcinoma) of the lung revealed the following overall general rates of mutation in lung adenocarcinoma at our institution: EGFR (exons 19 and 21), 10%; K-ras, 23%; and BRAF, 5.5% (S. Dacic et al, 2009, unpublished data).

The χ2 and Fisher exact tests were used for categorical data. Two-tailed P values of less than .05 were considered significant.

Results

The clinicohistopathologic and molecular profiles of the 15 cases of MPA are summarized in Table 1. The male/ female ratio was 8:7 (1.1), and ages ranged from 50 to 80 years (mean and median, 68 years). Of the 15 patients, 13 (87%) were 60 years or older at diagnosis. All patients currently smoked or had a history of smoking cigarettes. Tumor sizes ranged from 0.5 to 6.5 cm (mean, 3.4 cm; median, 3.0 cm), and 10 (67%) of the tumors were 3 cm or larger.

Image 1

Micropapillary lung adenocarcinoma. At low magnification, a nodular growth pattern with prominent desmoplastic stroma is shown (H&E, x40).

Image 2

Micropapillary lung adenocarcinoma. Dyscohesive papillary clusters of cytologically malignant cells “float” within air spaces and are focally associated with lepidic growth (H&E, x120).

Image 3

Micropapillary lung adenocarcinoma (MPA). While MPA frequently shows “hobnail” cytologic and bronchioloalveolar features, acinar and tubular architectures are common, as are columnar and polygonal cell cytologic features with abundant intracellular and extracellular mucin production (H&E, x100).

Micropapillary growth constituted 75% to 100% of the tumors Image 1 and Image 2. In 13 cases, secondary minor growth patterns included a lepidic pattern in 7 (54%), an acinar growth pattern in 5 (38%), and solid growth in 1 (8%); 2 cases showed micropapillary growth exclusively. Cytologically, the tumors were extremely heterogeneous, although we attempted to separate out nonmucinous hobnail and terminal reserve unit–type micropapillary adenocarcinoma (n = 4) from tumors having a predominant columnar or polygonal cell configuration (n = 8), with the remainder having mixed cell populations (n = 3) Image 3.7,17 Of the 4 with pure hobnail or terminal reserve differentiation, 2 had K-ras mutations, 1 an EGFR mutation, and 1 a BRAF mutation. Of the tumors with pure columnar or polygonal cell change, 2 had EGFR mutations, 1 a K-ras mutation, and 1 a BRAF mutation. Of the 15 adenocarcinomas, 5 showed clear cell differentiation, whereas only 1 showed signet-ring change. Intracellular and extracellular mucin production with diastase-predigested periodic acid–Schiff and mucicarmine stains was seen in 8 (53%) of 15 cases. Psammoma bodies were noted in 4 (27%) of 15 cases.

Of 7 cases associated with secondary lepidic growth, 6 demonstrated the following mutations: 3 (43%) of 7, K-ras; 2 (29%) of 7, EGFR; and 1 (14%) of 7, BRAF V600E. Among the MPAs showing mucin production with histochemical stains, 4 (50%) of 8 showed the following mutations: 2 (25%) of 8, BRAF V600E; 1 (13%) of 8, K-ras; and 1 (13%) of 8, EGFR.

Molecular studies demonstrated that 11 (73%) of 15 cases had mutations involving the 3 genes investigated: 5 (33%) showed K-ras mutations, 3 (20%) demonstrated EGFR mutations, and 3 (20%) showed BRAF V600E mutations Figure 1. Of the 3 EGFR mutations, 2 were deletions in exon 19 and 1 was a point mutation in exon 21. EGFR amplification was detected in 2 (13%) of 15 cases, with 1 of these 2 cases associated with EGFR mutations. All K-ras mutations were at codon 12 leading to substitution of glycine with phenylalanine (n = 1), cysteine (n = 3), or alanine (n = 1). Gly12Phe is a rare type of K-ras mutation in which 2 nucleotides are substituted at codon 12 (GGT to TTT) instead of the usual 1-nucleotide substitution. We attempted to correlate the mutations of these 3 genes with morphologic findings. No clear association of mutations with the morphologic or clinical features shown in Table 1 was noted (P > .05).

Discussion

Micropapillary growth in pulmonary adenocarcinomas reflects an aggressive subset of lung adenocarcinomas with a poor prognosis.2,3,5,6,810 In analyzing the relevant studies, one problem is achieving agreement on a definition for micropapillary adenocarcinoma; articles reporting on micro-papillary growth have indicated its presence in between 5% and 100% of their study populations.210 In our study of 15 cases of primary MPA, we used an original definition suggested by Silver and Askin13 that a papillary adenocarcinoma be composed of greater than or equal to 75% papillary growth. By using this definition for MPA, we confirmed the general observations of other studies having a less stringent definition. In particular, micropapillary adenocarcinoma is typically a malignancy of older people, with an equal sex distribution, that manifests more often at a late stage with intrapulmonary and extrapulmonary metastases. Our study is unique in focusing on the 3 major oncogenes reported in lung adenocarcinoma, although we recognize that between 30% and 70% of lung adenocarcinomas have complex genetic mutations not tied to EGFR, K-ras, and BRAF.18 Still, in our study group, 73% of the cases demonstrated mutations of K-ras (33%), EGFR (20%), and BRAF (20%), making MPA unusual in its frequency of involvement of these 3 genes compared with our institutional percentages in lung adenocarcinoma: EGFR, 10%; K-ras, 23%; and BRAF, 5.5% (see the “Materials and Methods” section). This percentage far outweighs the cumulative percentages of 40% usually associated with primary conventional lung adenocarcinoma.1820

Most studies on micropapillary growth and lung adenocarcinomas have focused on a higher than usual incidence of EGFR mutations in papillary and micropapillary adenocarcinomas of lung.7,11 Our study, in fact, confirmed that EGFR mutations are twice as common in MPA as reported in the Western literature. However, our study also highlights that EGFR is not the only gene associated with MPA, with K-ras and BRAF mutations present in more than 50% of our study population. Furthermore, while the association of EGFR mutations with an absence of smoking has been emphasized in Far Eastern and Western populations, our study also shows that EGFR mutations in MPA also occur in cigarette smokers.11,17,21,22 In our study, we defined a history of cigarette smoking in absolute terms—any patient with a history of cigarette smoking was considered as having a risk of developing adenocarcinoma, in contrast with other studies, such as a study by Motoi et al,11 who defined nonsmokers as patients having less than a 15-pack-year history of smoking.

EGFR mutations have also been reported primarily in adenocarcinomas having type II alveolar pneumocyte differentiation, variously reported as “hobnail configuration” or “terminal reserve unit” cytologic features.7,11,17 Although EGFR mutations were certainly seen in cells with these differentiation characteristics, EGFR mutations were also observed in MPAs with columnar and polygonal cell differentiation, whereas K-ras and BRAF mutations were seen in adenocarcinomas with hobnail or terminal reserve unit cell differentiation. Furthermore, although it has been emphasized that mucin-producing, goblet cell–type bronchioloalveolar adenocarcinomas are unassociated with EGFR mutations, our micropapillary adenocarcinomas produced mucin and had EGFR mutations, as well as K-ras and BRAF mutations.

Figure 1

Examples of mutations found in micropapillary lung adenocarcinoma. A, Direct nucleotide sequencing shows a 15-base-pair (bp) deletion in exon 19 of the EGFR gene. B, Point mutation (CTG to CGG) at codon 858 exon 21 of the EGFR gene detected by sequencing. C, K-ras codon 12 mutation (GGT to TGT) detected by sequencing leads to glycine to cysteine amino acid change. D, BRAF V600E mutation detected by real-time polymerase chain reaction (PCR) and post-PCR melting curve analysis demonstrates the mutant peak at a melting temperature of 58°C in addition to the wild-type (WT) peak.

The majority of our MPA cases demonstrated secondary lepidic and acinar growth patterns ranging from 1% to 25% of the tumor. There was no strong correlation between secondary growth pattern and molecular profile, although overall, K-ras mutations were more frequent in tumors with lepidic and papillary growth, whereas BRAF mutations were slightly more frequent in tumors with acinar growth. Nevertheless, in this group of MPAs, it is important to emphasize that in contrast with the findings of other studies, lepidic growth in these adenocarcinomas did not exclude EGFR mutations.11

To date, only 2 studies7,11 have looked specifically at the molecular alterations in MPAs, and neither used the definitions of Silver and Askin.13 In the study by Ninomiya et al,7 MPA was defined as an adenocarcinoma with 25% of the tumor having micropapillary growth. By using this definition, the authors showed that MPA closely resembled their EGFR+ adenocarcinomas in their demographics in this Japanese cohort: EGFR mutations in 4 exons occurred in 59% of cases; patients were usually nonsmoking women, and there was a strong association with a bronchioloalveolar and papillary growth pattern. Our study in a Western population, using more rigorous criteria for the definition of MPA, suggests that this tumor has a wider demographic and molecular spectrum. Motoi et al11 did not define MPA by absolute percentages of micropapillary growth, the result being the inclusion of conventional adenocarcinomas and papillary adenocarcinomas in the analysis, in which micropapillary architecture represented only the major component in relative terms but not necessarily in percentages greater than 75%. We confirmed their observation that EGFR mutations are more common in MPA than in conventional adenocarcinoma, but we also noted that MPA is present in smokers, displays mucinous differentiation, and is associated with K-ras and BRAF mutations, as noted by others.23,24

In the literature, EGFR mutations have largely been associated with bronchioloalveolar adenocarcinoma, invasive adenocarcinomas with prominent lepidic growth, and papillary adenocarcinoma.7,11,20,23,24 Micropapillary adenocarcinoma of the lung needs to be added to this group, although as emphasized before, mutations in this group are not restricted to EGFR, thus warranting a comprehensive molecular analysis should personalized therapies be contemplated. For example, BRAF mutations occur in approximately 5% of pulmonary adenocarcinomas, often with a papillary architecture, but in the micropapillary subgroup, the incidence rises to approximately 20%, similar to that seen in thyroid and ovarian papillary adenocarcinoma.16 It may be worthwhile to add this gene to a molecular panel when one sees this architectural pattern given that some personalized therapies for BRAF mutation (akin to tyrosine kinase inhibitor therapy in EGFR-mutated adenocarcinoma) have been identified.2530

This study represents the first attempt to rigorously evaluate the molecular alterations in MPAs by using a uniform definition and to correlate this growth pattern with K-ras, EGFR, and BRAF mutations. We conclude that K-ras, EGFR, and BRAF mutations occur at an increased frequency in lung adenocarcinomas showing greater than 75% micropapillary growth and that these mutations are seen in smokers and in adenocarcinomas with mucin production. K-ras mutations certainly predominate, but BRAF and EGFR mutations occur at a higher incidence than in conventional lung adenocarcinomas reported in the literature and in our institutional experience.

Acknowledgments

We thank Kathy Cieply for fluorescence in situ hybridization studies and Linda Shab and Tom Bauer for photographic aid.

References

  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.
View Abstract