Lung cancer is the most common cause of cancer-related death worldwide.1 Adenosquamous lung carcinoma (AdSqLC) is a subtype of non–small-cell lung cancer (NSCLC), accounting for 1.6% to 4.5% of lung cancers.2–4 According to World Health Organization classification, AdSqLC is defined as a carcinoma showing components of both adenocarcinoma and squamous cell carcinoma, with each component comprising at least 10% of the tumor.5 AdSqLCs have more aggressive behavior and worse prognosis than adenocarcinoma or squamous cell carcinoma.6,7
One of the most important and promising treatment strategies for NSCLC involves the subdivision of tumors into clinically relevant molecular subsets, using a classification based on the presence or absence of known so-called driver mutations, including EGFR, ERBB2, KRAS, BRAF, PIK3CA, AKT1, RET, and ALK.8,9 EGFR tyrosine kinase inhibitors are currently recommended as the first-line option for patients with EGFR-sensitizing mutations in the metastatic setting. In addition and more recently, crizotinib has been recommended as first-line systemic therapy for patients with advanced ALK-rearranged NSCLC.10 Mutations in ERBB2, PIK3CA, BRAF, RET, MEK1, and AKT1 also define subsets of NSCLCs that may be amenable to treatment with other specific kinase inhibitors.8,9 A few studies with small sample size found that the frequency of EGFR mutation in AdSqLCs ranged from 15% to 44% in East Asian population.11–14 However, whether AdSqLCs harbor other mutations and details of their prevalence are still largely unknown. With an estimated 1,608,800 new cases of lung cancer per year worldwide in 2005, this equates to more than 25,000 new adenosquamous lung tumors per year,1 thus clarifying that mutation status could have therapeutic implications.
Because the understanding of genetic bases of AdSqLC is limited, there is no standard treatment currently available for the disease. AdSqLCs had a relatively high frequency of EGFR mutation,11–14 which is one of characteristics of lung adenocarcinoma and thereby most clinicians are prone to treat them as lung adenocarcinoma. By microdissection, previous studies found that AdSqLC is a heterogeneous disease with two cell components each of which carried the same EGFR mutations, suggesting that the tumors are relatively clonal.11,13 In addition, a recent study reported that a female AdSqLC patient with EGFR activating mutation treated with gefitinib showed a complete response lasting 3 years.15 Thus, it is hypothesized that AdSqLC and adenocarcinoma originate from common cancer stem cells. Poorly differentiated lung adenocarcinomas also showed very aggressive clinical course than moderate/well-differentiated adenocarcinomas.16,17 However, it remains unclear whether AdSqLCs have mutation profiles and clinicopathologic characteristics similar to poorly differentiated adenocarcinomas.
In most AdSqLCs, glandular component shows acinar, lepidic, micropapillary, or papillary growth pattern.18 It is also noted that some of tumors showed solid growth pattern in glandular component.4 Previous studies demonstrated that majority of ALK-rearranged lung adenocarcinomas showed a solid growth pattern. In contrast, only a small percentage of ALK wild-type tumors showed solid growth.19,20 In our previous work, we also found that 73% of RET-rearranged lung adenocarcinomas showed solid pattern.21 A squamous cell carcinoma marker, TP63, was expressed in a majority of ALK-rearranged adenocarcinomas. However, staining of TP63 was seen in a minority of adenocarcinomas (9%–32%), suggesting that its expression in ALK-rearranged tumor might not be simply aberrant.20,22,23 The p63 transcription factor (TP63) plays a critical role in squamous cell carcinoma, which is a master regulator of squamous cell differentiation.22 It is interesting to note that some of AdSqLCs can show both solid and squamous component (TP63-positive). Its mutational characteristics are unknown.
Here we examined the mutational status of EGFR, KRAS, ERBB2, BRAF, PIK3CA, AKT1, RET, and ALK in a cohort of 76 Chinese patients with resected AdSqLCs. We also compared the clinicopathologic and mutational characteristics between 76 AdSqLCs and 646 lung adenocarcinomas.
MATERIALS AND METHODS
During the period from October 2002 to June 2012, an overall of 2195 patients were treated surgically for primary lung cancer in the Department of Thoracic Surgery of Fudan University Shanghai Cancer Center. Of these, we identified 76 (3.4%) with AdSqLC. Fifty-five patients had both frozen and corresponding formalin-fixed paraffin-embedded (FFPE) tumor tissues. For 21 patients diagnosed before October 2007, only FFPE tissues were available.
From October 2007 to August 2011, a total of 646 patients with lung adenocarcinoma were also operated on for a primary tumor, including 23 patients (3.5%) who during surgery were found unexpectedly to have pleural metastasis. For these cases, only biopsy samples of the metastatic sites were collected.
A consent form was signed by every patient or legal representative. The study was approved by the Committee for Ethical Review of Research at Shanghai Cancer Hospital. Medical records of all patients were reviewed to extract data on clinicopathologic characteristics, including sex, age, smoking history, stage, and treatment history. Patients were followed up in clinic or by telephone for disease recurrence and survival.
Tumor Pathology and Mutational Analyses
Each case was reviewed and both squamous and glandular components in at least 10% of the tumor by light microscope were confirmed by two pathologists (Drs. Li and Shen) after histologic diagnosis based on World Health Organization criteria.5 Squamous component showing identifiable keratinization, pearl formation, and/or intercellular bridges is required for the diagnosis. Histological subtypes of glandular component were determined by the new International Association for the Study of Lung Cancer/ATS/ERS multidisciplinary classification.21 Tumors showing acinar, lepidic, micropapillary, or papillary growth pattern glandular component were diagnosed as classical AdSqLC. The proportion of glandular and squamous components in classical AdSqLC was identified by microscopic examination. The diagnosis of AdSqLC showing solid growth pattern in glandular component was also verified by immunohistochemical biomarkers (TTF-1, CK7, Napsin A, CK5/6, and/or p63).
The mutational status of EGFR (exons18–21), ERBB2 (exon 20), KRAS (exons 2–3), BRAF (exons 11–15), and AKT1 (exon 4) was determined using polymerase chain reaction–based direct sequencing and verified by DNA sequencing analysis. For detection of ALK and RET fusions, primers were designed to amplify all known fusion variants with the use of cDNA (Supplementary Table 1, Supplemental Digital Content 1, http://links.lww.com/JTO/A453). All mutations were verified by analysis of an independent polymerase chain reaction isolate.
To study further the glandular and squamous components of tumors by microdissection, clearly distinguishable AdSqLCs were next studied, which included the following criteria: tumors harboring a mutation; tumor specimen containing a minimum of 50% tumor cells to ensure adequate material; and corresponding normal tissue also being available for analysis. To avoid the contamination of one cell component by the other, samples with two cell components that intermingled on histologic examination were excluded.
Fluorescence In Situ Hybridization
As previously described,24 ALK rearrangements were detected in FFPE specimens with the use of a commercially available break-apart probe for the ALK gene (Vysis LSI ALK Dual Color; Abbott Molecular, Abbott Park, IL).
Data were analyzed using the Statistical Package for the Social Sciences Version 16.0 Software (SPSS Inc., Chicago, IL) or Prism 5.0 (Graph Pad Software Inc., La Jolla, CA). Relapse-free survival (RFS) and overall survival (OS) data of AdSqLC patients diagnosed from June 2007 to August 2011 were included and compared with data from lung adenocarcinoma patients diagnosed from April 2008 to April 2010. The Kaplan–Meier method was used to estimate RFS and OS, and the differences were compared using the log-rank test. The two-sided significance level was set at a p value less than 0.05.
A total of 76 AdSqLCs from a Chinese population were studied. All cases were rereviewed by the two pathologists for confirmation of tumor histology. The characteristics of these AdSqLCs are listed in Supplementary Table 2 (Supplemental Digital Content 2, http://links.lww.com/JTO/A454). There are more smokers in AdSqLC than in adenocarcinoma (p < 0.0001).
Spectrum of Mutations
Tumor specimens containing both components were studied in mutational analysis; 56.6% (43 of 76) harbored known mutant kinases. Mutations in the EGFR kinase domain were observed in 31.6% tumors (24 of 76). One tumor harbored both exon 19 deletion and PIK3CA mutations. The remaining 19 tumor-carrying mutations included eight with KRAS mutation, four with ALK fusion, two with AKT1 mutation, one with ERBB2 mutation, three with KIF5B-RET fusion, and one with PIK3CA mutation. No mutation was found in BRAF. Of the samples 43.4% (33 of 76) did not harbor mutations in EGFR, KRAS, ALK, ERBB2, PIK3CA, AKT1, RET, or BRAF (Fig. 1).
There was a significantly higher rate of EGFR mutations in never smokers (51.5% versus 16.3 %; p = 0.001). EGFR mutations were found more frequently in tumors larger than 3 cm (41.3% versus 16.7%; p = 0.024; Supplementary Table 3, Supplemental Digital Content 3, http://links.lww.com/JTO/A455). The clinicopathologic characteristics of individual patients with EGFR/ERBB2/KRAS/BRAF/PIK3CA/AKT1 mutations and ALK/RET fusions are summarized in Supplementary Table 4 (Supplemental Digital Content 4, http://links.lww.com/JTO/A456).
Of 24 EGFR kinase domain mutations, there were two exon 20 insertion mutations and one T790M mutation, which can confer resistance to gefitinib/erlotinib treatment. Six of eight KRAS mutations (75%) were found in smokers and 87.5% of patients with KRAS mutation were men. We also found that three of four ALK fusions (75%) were found in smokers.
Mutational and Clinicopathologic Characteristics of AdSqLCs
Fifty-three of 76 AdSqLCs (69.7%) showing acinar, lepidic, micropapillary, or papillary glandular growth patterns were classified as classical AdSqLCs. Representative microscopic images of histologic subtypes of glandular component in AdSqLCs are shown in Supplementary Figure 1 (Supplemental Digital Content 5, http://links.lww.com/JTO/A4570). Of the patients with classical AdSqLC, 69.8% (37 of 53) were men and 51% (27 of 53) were smokers. The proportion of two cell components was available in 48 samples. The frequency of EGFR mutations in glandular-dominant AdSqLCs (glandular component ≥50%) was significantly higher than that in squamous-dominant tumors (squamous component ≥50%; 80% versus 24.2%; p < 0.0001), whereas KRAS, AKT1, and PIK3CA mutations were all found in squamous-predominant AdSqLCs (Fig. 2A).
We further found that driver mutation profiles were quite similar between classical AdSqLCs and poorly differentiated lung adenocarcinomas (Fig. 2). There were no significant differences between classical AdSqLCs and poorly differentiated lung adenocarcinomas regarding frequency of EGFR or KRAS mutations (Table 1). We also found that the frequency of EGFR mutation in AdSqLCs was significantly lower than in adenocarcinomas (61.6% versus 31.6%; p < 0.0001).
Twenty-three of 76 AdSqLCs (30.3%) showed solid growth pattern in glandular component (Table 2). The frequency of fusion genes (RET and ALK) was surprisingly higher in AdSqLCs with solid growth glandular component than that with classical glandular component (30.4% versus 0%). The mutation rate of EGFR for AdSqLC with solid growth glandular component was only 4.3%, lower than that for classical AdSqLC (41.3%; p = 0.001; Table 1).
Mutational Analysis in Microdissected Classical AdSqLCs
Thirty-two of 53 classical AdSqLCs (60.4%) harbored mutations in one of the tested genes, and 16 (58.6%) of these were excluded from microdissection analysis. These included 10 samples with tumor content less than 50%, three with two cell components intermingled on histologic examination, and three with insufficient tissue. It was possible to perform microdissection on 13 tumors, including eight with EGFR mutations (1 case with an EGFR synonymous mutation, T790T), three with KRAS mutations, one with double mutations (EGFR and PIK3CA), and one with ERBB2 mutation. In nine of the 13, same mutation was found in both histologic components, whereas in four cases, mutations were found only in the glandular component, including two KRAS mutations (cases 4 and 30), one HER2 mutation (case 36; Fig. 3), and one synonymous EGFR mutation (G2370A; T790T).
The survival data were available in 58 AdSqLCs and 246 lung adenocarcinomas. Twenty-nine of 58 AdSqLC patients (50%) died during the follow-up. After surgery, nine of 22 patients (41%) with EGFR mutations are still relapse-free and 13 (59%) experienced a relapse. There were no significant differences in RFS or OS for AdSqLC patients with EGFR mutations compared with those with KRAS, ALK, or wild-type (Supplementary Figure 1, Supplemental Digital Content 5, http://links.lww.com/JTO/A4570). The 2-year OS for patients with EGFR mutations was 51.3% (95% confidence interval, 32%–83.6%). Patients with ALK rearrangements had a longer OS as compared with patients with KRAS mutations (p = 0.027). The number of AKT1, ERBB2, RET, and PIK3CA mutations were too small for further analysis.
We further compared the RFS and OS for patients with AdSqLCs and adenocarcinomas. There were no significant differences in RFS among patients with classical AdSqLC, poorly differentiated lung adenocarcinomas, and AdSqLCs with solid growth glandular component, but they all showed a worse OS compared with moderately to well-differentiated adenocarcinomas (Fig. 4).
It is becoming evident that subsets of NSCLC can be further defined at the molecular level by specific driver mutations that play a crucial role in tumor transformation. Such mutations have been found to occur in EGFR, ERBB2, KRAS, BRAF, PIK3CA, AKT1, RET, and ALK. These mutations have been widely studied mainly in lung adenocarcinomas. EGFR, ERBB2, KRAS, BRAF, and ALK mutations are much more frequent in adenocarcinomas than in squamous cell carcinomas,21 whereas PIK3CA mutations occur more frequently in squamous cell carcinomas25 and AKT1 mutations have been identified only in squamous cell carcinomas.26 A few studies have reported the frequency of EGFR and KRAS mutations in a limited number of AdSqLCs and that ALK fusions were also found incidentally, but did not note the status of other driver mutations.11–14,27 Here we delineate, for the first time, profiles of the major known driver mutations and their clinicopathologic characteristics in a large cohort of AdSqLCs in Chinese patients.
Prior studies found that EGFR mutation rate for AdSqLCs ranged from 15.4% to 44%, higher than that for squamous cell carcinoma, but less for adenocarcinoma.11,13,14,27 Therefore, it seems reasonable to assume that AdSqLC is a mix of various proportions of adenocarcinoma and squamous cell carcinomas. Our study showed that the frequency of EGFR mutation in AdSqLCs was significantly lower than in adenocarcinomas, which was consistent with prior studies.11,13,14,27 However, other gene mutations (KRAS, ALK, RET, PIK3CA, and AKT1) in AdSqLCs showed frequency comparable to that in adenocarcinomas. Our results can help explain the inconsistency of mutation profiles between adenocarcinoma and AdSqLC. We found that clinicopathologic and mutational characteristics of patients with classical AdSqLC were strikingly similar to that of poorly differentiated adenocarcinomas.
The similar characteristics indicate that the two subtypes may share a common mechanism of pathogenesis. In our study, with the use of strict criteria for confident microdissection, 12 of 29 classical AdSqLCs with mutations (41.4%) were included for microdissection analysis. Similar to previous results,11,13 all activating EGFR mutations and one KRAS mutation were present in both glandular and squamous components. These findings indicate that both adenocarcinomas and AdSqLCs may arise from undifferentiated adenocarcinoma stem cells, which are capable of differentiating into adenocarcinomas and a few of them have the potential to give rise to both adeno- and squamous cell components. AdSqLCs might still keep their undifferentiated cancer stem cells or poorly differentiated characteristics. However, we still have few clues as to the pathogenesis of AdSqLCs; further studies on transformation mechanisms of AdSqLCs are needed.
Patients with classical AdSqLCs shared an EGFR mutation rate similar to that of poorly differentiated adenocarcinoma. Because glandular component in classical AdSqLC is morphologically so different from that of squamous component, the relative proportion of glandular and squamous components can be easily identified. Thereby we found that 90% EGFR mutations (18 of 20) presented in tumors with a glandular component higher than 20%, rarely in squamous predominant (>70%) or AdSqLCs. We also found that ALK and RET fusions were identified in 30.4% of AdSqLCs with solid growth in glandular component, whereas EGFR mutations were only found in 4.3%. These findings suggested that the identification of subtypes and proportion of glandular component in AdSqLC would allow clinicians to select patients who are most likely to benefit from targeted therapy.
To date there is little information concerning KRAS/ERBB2 mutation-driven AdSqLC. Only seven KRAS mutations were found in prior studies (3 in Western and 4 in Asian populations) and no ERBB2 insertion mutation was identified in AdSqLCs.11–13,27 In our study, we identified eight AdSqLCs that harbored KRAS mutation. Of these, three were identified as solid-type AdSqLC, and five were classified as classical AdSqLC. Three of them with KRAS mutation (2 with lepidic and 1 with papillary pattern in glandular component) were microdissected. It was very interesting that KRAS mutation only presented in glandular component, not in the squamous component in two tumors with lepidic growth pattern, and both components harbored KRAS mutation in another tumor with acinar component. KRAS-driven lepidic component (formerly known as bronchioloalveolar carcinoma) is a carcinoma in situ, whereas squamous cell component without KRAS mutation is an invasive carcinoma. It seems plausible that they might result from a collision of adenocarcinoma and squamous cell carcinoma. However, both ERBB2-driven glandular component and ERBB2-negative squamous component were invasive carcinomas, which may result from subclonal heterogeneity. The two components might derive from common ERBB2-negative cancer stem cells, and subclones developing ERBB2 mutation form glandular components whereas cells without ERBB2 mutation evolve into squamous carcinomas. Prior study demonstrated that heterogeneous EGFR mutations in adenocarcinoma were because of heterogeneous amplification of a mutated allele and low sensitivity of direct sequencing.28 Although only classical AdSqLCs with tumor cell content higher than 50% were selected for microdissection, we do not exclude the possibility that heterogeneous distribution of KRAS and ERBB2 mutations in AdSqLCs may result from technical artifact of direct sequencing.
Finally, our results demonstrated that classical AdSqLC shared clinicopathologic and mutational characteristics very similar to those of poorly differentiated adenocarcinoma. We also found that classical AdSqLC patients had a worse OS compared with patients with poorly differentiated adenocarcinoma, indicating that relapsed classical AdSqLCs may have more aggressive behavior and respond poorly to current therapies. However, studies with a much larger sample size and longer duration of follow-up are still necessary to confirm these results.
Supported by Grant No. 81101761, 81172218 from the National Natural Science Foundation of China; Key Construction Program of the National “985” Project (985-YFX0102); Grants No. WJP1101 from Wu Jieping Foundation; Grant from Shanghai Municipal Health Bureau; Grant No. SHDC12012308 from Shanghai Hospital Development Center. We also thank Dr. Matthew Meyerson for his valuable advice regarding this study.
1. Parkin DM, Bray F, Ferlay J, Pisani P. Global cancer statistics, 2002. CA Cancer J Clin. 2005;55:74–108
2. Filosso PL, Ruffini E, Asioli S, et al. Adenosquamous lung carcinomas: a histologic subtype with poor prognosis. Lung Cancer. 2011;74:25–29
3. Uramoto H, Yamada S, Hanagiri T. Clinicopathological characteristics of resected adenosquamous cell carcinoma of the lung: risk of coexistent double cancer. J Cardiothorac Surg. 2010;5:92
4. Gawrychowski J, Bruliński K, Malinowski E, Papla B. Prognosis and survival after radical resection of primary adenosquamous lung carcinoma. Eur J Cardiothorac Surg. 2005;27:686–692
5. Travis WD, Colby TV, Corrin B, et al.In Collaboration with L.H. Sobin and Pathologists from 14 Countries. Histological Typing of Lung and Pleural Tumours (International Histological Classification of Tumours). 1999 Geneva World Health Organization:1–55
6. Cooke DT, Nguyen DV, Yang Y, Chen SL, Yu C, Calhoun RF. Survival comparison of adenosquamous, squamous cell, and adenocarcinoma of the lung after lobectomy. Ann Thorac Surg. 2010;90:943–948
7. Maeda H, Matsumura A, Kawabata T, et al.Japan National Hospital Organization Study Group for Lung Cancer. Adenosquamous carcinoma of the lung: surgical results as compared with squamous cell and adenocarcinoma cases. Eur J Cardiothorac Surg. 2012;41:357–361
8. Pao W, Girard N. New driver mutations in non-small-cell lung cancer. Lancet Oncol. 2011;12:175–180
9. Kohno T, Ichikawa H, Totoki Y, et al. KIF5B-RET fusions in lung adenocarcinoma. Nat Med. 2012;18:375–377
10. National Comprehensive Cancer Network. NCCN guidelines: nonsmall cell lung cancer.Version 3.2011. 2011 Fort Washington, PA National Comprehensive Cancer Network
11. Kang SM, Kang HJ, Shin JH, et al. Identical epidermal growth factor receptor mutations in adenocarcinomatous and squamous cell carcinomatous components of adenosquamous carcinoma of the lung. Cancer. 2007;109:581–587
12. Toyooka S, Yatabe Y, Tokumo M, et al. Mutations of epidermal growth factor receptor and K-ras genes in adenosquamous carcinoma of the lung. Int J Cancer. 2006;118:1588–1590
13. Jia XL, Chen G. EGFR and KRAS mutations in Chinese patients with adenosquamous carcinoma of the lung. Lung Cancer. 2011;74:396–400
14. Sasaki H, Endo K, Yukiue H, Kobayashi Y, Yano M, Fujii Y. Mutation of epidermal growth factor receptor gene in adenosquamous carcinoma of the lung. Lung Cancer. 2007;55:129–130
15. Iwanaga K, Sueoka-Aragane N, Nakamura T, Mori D, Kimura S. The long-term survival of a patient with adenosquamous lung carcinoma harboring EGFR-activating mutations who was treated with gefitinib. Intern Med. 2012;51:2771–2774
16. Chapelier A, Fadel E, Macchiarini P, et al. Factors affecting long-term survival after en-bloc resection of lung cancer invading the chest wall. Eur J Cardiothorac Surg. 2000;18:513–518
17. Li C, Sun Y, Fang Z, et al. Comprehensive analysis of epidermal growth factor receptor gene status in lung adenocarcinoma. J Thorac Oncol. 2011;6:1016–1021
18. Bastide K, Ugolin N, Levalois C, Bernaudin JF, Chevillard S. Are adenosquamous lung carcinomas a simple mix of adenocarcinomas and squamous cell carcinomas, or more complex at the molecular level? Lung Cancer. 2010;68:1–9
19. Yoshida A, Tsuta K, Watanabe S, et al. Frequent ALK rearrangement and TTF-1/p63 co-expression in lung adenocarcinoma with signet-ring cell component. Lung Cancer. 2011;72:309–315
20. Yoshida A, Tsuta K, Nakamura H, et al. Comprehensive histologic analysis of ALK-rearranged lung carcinomas. Am J Surg Pathol. 2011;35:1226–1234
21. Wang R, Hu H, Pan Y, et al. RET fusions define a unique molecular and clinicopathologic subtype of non-small-cell lung cancer. J Clin Oncol. 2012;30:4352–4359
22. Rekhtman N, Ang DC, Sima CS, Travis WD, Moreira AL. Immunohistochemical algorithm for differentiation of lung adenocarcinoma and squamous cell carcinoma based on large series of whole-tissue sections with validation in small specimens. Mod Pathol. 2011;24:1348–1359
23. Kim MJ, Shin HC, Shin KC, Ro JY. Best immunohistochemical panel in distinguishing adenocarcinoma from squamous cell carcinoma of lung: tissue microarray assay in resected lung cancer specimens. Ann Diagn Pathol. 2013;17:85–90
24. Wang R, Pan Y, Li C, et al. The use of quantitative real-time reverse transcriptase PCR for 5′ and 3′ portions of ALK transcripts to detect ALK rearrangements in lung cancers. Clin Cancer Res. 2012;18:4725–4732
25. Lee SY, Kim MJ, Jin G, et al. Somatic mutations in epidermal growth factor receptor signaling pathway genes in non-small cell lung cancers. J Thorac Oncol. 2010;5:1734–1740
26. Do H, Salemi R, Murone C, Mitchell PL, Dobrovic A. Rarity of AKT1 and AKT3 E17K mutations in squamous cell carcinoma of lung. Cell Cycle. 2010;9:4411–4412
27. Tochigi N, Dacic S, Nikiforova M, Cieply KM, Yousem SA. Adenosquamous carcinoma of the lung: a microdissection study of KRAS and EGFR mutational and amplification status in a western patient population. Am J Clin Pathol. 2011;135:783–789
28. Yatabe Y, Matsuo K, Mitsudomi T. Heterogeneous distribution of EGFR mutations is extremely rare in lung adenocarcinoma. J Clin Oncol. 2011;29:2972–2977
Adenosquamous lung carcinoma; Mutations; EGFR
Copyright © 2014 by the European Lung Cancer Conference and the International Association for the Study of Lung Cancer.