Abstract
Basaloid carcinoma represents a rare variant of nonsmall cell lung cancer (NSCLC), which has shown a poor prognosis in a number of studies. Although it is considered to derive from a pluri- or multipotent pulmonary stem cells, little is known about the expression and clinical significance of stem cell antigens in this variant.
Stage-specific embryonic antigen-4 (SSEA-4) was analysed by immunohistochemistry in 38 patients with resected early-stage basaloid NSCLC who had a median follow-up of 72.9 months. The expression of SSEA-4 was related to clinico-pathological characteristics, to the expression of the adult stem cell antigens CD117, CD133 and breast cancer resistance protein 1 (BCRP1), and to prognosis.
SSEA-4 was positive in 37% of the specimens and showed no association with clinico-pathological characteristics or the expression of adult stem cell antigens. Cox proportional hazards regression analysis revealed a 6.0-fold increased risk of relapse (p = 0.001) and a 4.2-fold increased risk of disease-related mortality (p = 0.017) in SSEA-4-positive patients, while SSEA-4-negative patients showed a prognosis comparable with that of other early-stage NSCLC.
SSEA-4 is expressed in a fraction of basaloid NSCLC and is associated with poor prognosis.
Basaloid lung cancer is a rare variant of pulmonary squamous cell carcinoma (SCC) or large cell carcinoma and accounts for 4.8–6.3% of all nonsmall cell lung cancer (NSCLC) cases [1, 2]. In contrast to overall NSCLC, which has a 5-yr survival rate of, on average, 40–60% and a median survival time of ∼50 months in stage I/II disease, analyses of two European collectives comprising 38 and 90 patients with basaloid NSCLC revealed a significantly lower 5-yr survival rate of 10–27% and a shorter median survival time of 20–29 months [1, 3–6]. The basal cell phenotype of these carcinomas and their co-occurrence with virtually all histological subtypes of NSCLC soon led to the assumption that this variant derives from a pluri- or multipotent pulmonary stem cell [1, 5, 7]. In recent years, several in vitro and in vivo studies indicated the presence of stem-like cells in NSCLC. Those were characterised by: the expression of stem cell antigens, such as CD117, CD133 and/or breast cancer resistance protein 1 (BCRP1); the ability to regenerate the primary tumour; an increased metastatic capability; and drug resistance [8–10]. Previous studies in early-stage disease indicated that expression of CD133 and BRCP1 predicts an increased risk of relapse and disease-related mortality after pulmonary resection [11, 12]. In contrast to adult stem cell antigens, the knowledge about the expression and significance of embryonic stem cell (ESC) antigens in lung cancer is scant. The markers most frequently used to identify ESC are the glycolipid carbohydrate epitopes stage-specific embryonic antigen (SSEA)-3 and -4 and the transcription factor Oct-4 [13]. Recent studies demonstrated expression of Oct-4 in a subset of NSCLC cells with stem cell properties and an unfavourable prognostic significance of this antigen in adenocarcinomas with lepidic growth pattern [14–17]. In contrast, SSEA-4, which characterises very early ESCs, has yet been analysed in teratocarcinomas, testicular germ cell tumours and a single study of ovarian cancer, but not in lung cancer [18–20]. Analyses of the expression of stem cell antigens are lacking so far, particularly in basaloid NSCLC, which is predominantly suggested to have an association with pulmonary stem cells. In our study, we analysed tumour tissue of 38 completely resected stage I/II patients with basaloid NSCLC for the expression and prognostic significance of SSEA-4. Moreover, the results were related to clinico-pathological characteristics and the expression of the adult stem cell antigens CD117, CD133 and BCRP1.
MATERIALS AND METHODS
Patient and sample characteristics
A total of 38 previously untreated patients with basaloid NSCLC who underwent complete pulmonary resection between 2002 and 2009 at the Dept of Thoracic Surgery, Thoraxklinik, University of Heidelberg, Heidelberg, Germany, were analysed. Tissue specimens and follow-up data were obtained from the tissue bank and the lung cancer registry of the Thoraxklinik and the tissue bank of the National Center for Tumour Diseases (NCT), University of Heidelberg. Patients had given informed consent following the 2000 revision of the guidelines of the Declaration of Helsinki and the local Ethics Committee of the Medical Faculty Heidelberg. Pre-operative and follow-up assessments were performed according to the guidelines of the German Respiratory Society last published in 2010 [21]. The stage was determined according to the seventh edition of the TNM classification of malignant tumours [3]. Detailed patient characteristics are given in table 1.
Histological classification
The histological classification of a basaloid NSCLC variant was based on the criteria published in the 2004 revision of the World Health Organization classification of tumours and included: 1) solid lobular or anastomotic trabecular pattern growing invasively in a finger-like fashion from the bronchial and/or glandular duct lining; 2) small cuboidal to fusiform cells of 12–15 μm mean diameter with scant, but visible cytoplasm, and moderately hyperchromatic nuclei without nuclear moulding or prominent nucleoli; 3) peripheral palisading with radially arranged cells at the periphery of the tumour lobules; and 4) a mitotic index of ≥15–44 mitoses per 10 high-power fields [22]. Haematoxylin-stained tissue sections were used for histological classification and grading. The diagnosis of a squamous variant was based on the presence of intercellular bridges or individual cell keratinisation within the basal cell component in <50% of the tumour area. Additionally, the expression of the neuroendocrine (NE) markers neuron-specific enolase, synaptophysin, chromogranin A, CD56 and the high molecular weight cytokeratins (HMWCKs) 1, 5, 10 and 14 was analysed to exclude large cell neuroendocrine or small cell carcinomas. The diagnosis of a basaloid carcinoma was maintained when HMWCKs were expressed and clear-cut staining for at least one specific NE marker was lacking.
Immunohistochemical and immunocytochemical analyses
Tissue microarrays (TMA) of formalin-fixed, paraffin-embedded tissue derived from pulmonary resections were prepared by the tissue bank of the NCT, as previously described [23]. For each patient, two 1.2-mm diameter tissue cores each of the tumour centre, the invasion front and histologically normal lung tissue were spotted. In half of the cases, additionally, whole tissue sections were prepared. 2-μm sections of the TMA or the whole tissue were deparaffinised with xylene and rehydrated in graded alcohol series. For antigen retrieval, the sections were boiled in target retrieval buffer pH 6 (Dako, Glostrup, Denmark) for 15 min. Sections for CD133 and BCRP1 staining were additionally blocked by avidin–biotin treatment. Subsequent steps were performed in an immunostaining device (Autostainer; Dako). Briefly, the sections were incubated with the primary antibody for 30 min, washed with PBS/Tween 20, incubated with the secondary antibody for 20 min and washed again. Endogenous peroxidase was blocked by incubation with peroxidase-blocking solution (Dako) for 5 min. Detection was based on the avidin–biotin peroxidase principle using AEC as chromogen (Dako REAL Detection System Peroxidase/AEC, Rabbit/Mouse). The sections were counterstained with Mayer's haematoxylin for 5 min and mounted with coverslips in Aquatex mounting medium (Merck KGaA, Darmstadt, Germany). The following primary antibodies, clones and dilutions were used: SSEA-4 (clone MC-813-70, 1:100; Millipore, Billerica, MA, USA), CD117 (polyclonal, 1:50, Dako), CD133 (polyclonal (ab19898), 1:100; Abcam plc, Cambridge, MA, USA), BCRP1 (clone BXP-21, 1:100; Abcam), Ki67 (clone MiB1, 1:200; Dako), CD56 (clone 1B6, 1:50; Novocastra, A. Menarini Diagnostics Deutschland, Berlin, Germany), synaptophysin (clone 27G12, 1:400; Novocastra), neuron-specific enolase (clone MIG-N3, 1:10,000; DCS Innovative Diagnostik-Systeme Dr. Christian Sartori GmbH & Co. KG, Hamburg, Germany), synaptophysin (clone LK2H10, 1:10; Abcam) and HMWCK (clone 34βE12, 1:100, Abcam) recognising cytokeratin (CK) 1, 5, 10 and 14. For negative controls, the primary antibody was omitted. Teratocarcinoma tissue derived from the tissue bank of the NCT was used as a positive control for SSEA-4. For analysis of SSEA-4 expression in nontumoral conditions, lung tissue of healthy subjects with spontaneous pneumothorax was used and provided by the tissue bank of the NCT. The analysis was performed by two independent observers (E. Herpel and P.A. Schnabel). Samples were considered positive, if either dispersed expression or at least one focus with distinct staining was present. The TMA were scanned at 400× magnification using Aperio ImageScope v10.1.3.2028 software (Aperio Technologies Inc., Vista, CA, USA).
The human embryonic carcinoma cell line NTERA-2 (Deutsche Sammlung für Mikroorganismen und Zellkulturen GmbH, Braunschweig, Germany) and the basaloid NSCLC cell line 2427T were cultured as indicated by the supplier or as previously published, respectively [24, 25]. For immunocytochemical analyses, cytospins of each 1 million NTERA-2 or 2427T cells were prepared in a Shandon CytoSpin III cytocentrifuge (Fisher Scientific GmbH, Schwerte, Germany). The cells were fixed with neutral buffered 10% formalin for 5 min, boiled in target retrieval buffer pH 6 (Dako) for 15 min and permeabilised with Triton X-100 for 10 min. All subsequent steps were performed as described above.
Statistical analyses
The follow-up was defined as the Kaplan–Meier estimate with reversed status indicator. Death censored the true but unknown observation time, censoring was interpreted as end-point. Thus, the unobservable follow-up time of a deceased patient was interpreted as the follow-up time that potentially would have been obtained if the patient had not died. Survival time was determined from the date of first diagnosis until last follow-up or reported death. Nondisease-related death was censored. The disease-free survival was determined from the date of first diagnosis until diagnosis of relapse or disease-related death. Survival times were analysed using the Kaplan–Meier method and the log-rank test. For multivariate analysis the Chi-squared test and Wilcoxon rank sum test were used to evaluate the difference between groups. A p-value <0.05 was considered statistically significant. The statistical analyses were performed using SAS® version 9.2 (SAS Institute, Cary, NC, USA).
RESULTS
Expression of SSEA-4 in basaloid NSCLC
Tissue specimens of 38 patients with stage I/II basaloid NSCLC were analysed for the expression of SSEA-4. The collective contained 24 pure basaloid carcinomas, 13 basaloid carcinomas with a SCC component and one basaloid carcinoma with an adenocarcinoma component. In all specimens, the basaloid component comprised ≥60% of the tumour bulk. SSEA-4 showed diffuse membranous/cytoplasmic expression in 14 (37%) out of 38 of the specimens (figs 1a and 2a). The positive control, a teratocarcinoma, likewise showed diffuse cytoplasmic/membranous and, in some cells, also showed nuclear staining (fig. 2b). In the NSCLC specimens, positive staining was restricted to the basaloid component, while the well differentiated adenocarcinoma (AC) or SCC components were negative for SSEA-4 (fig. 2c and 2d). The alveolar epithelium of patients and healthy subjects was negative for SSEA-4 (figs 1i and 2e), whereas the tracheobronchial epithelium displayed membranous/cytoplasmic staining of the basal cells and cytoplasmic accumulation at the apical cell pole of ciliated cells (figs 1e and 2f). All specimens displayed reactivity for HMWCK, while clear-cut expression of specific NE markers was absent (fig. 3). Remarkably, the basaloid NSCLC cell line 2427T which shows properties related to stem cells, i.e. high tumorigenicity in the animal model and reassembling of the original tumour histomorphology, also displayed expression of SSEA-4. Similarly, the positive control, the embryonic carcinoma cell line NTERA-2 showed distinct membranous/cytoplasmic staining (fig. 4).
SSEA-4, clinico-pathological characteristics and adult stem cell antigens
The expression of SSEA-4 showed no association with clinico-pathological characteristics or the expression of adult stem cell antigens (tables 1 and 2). CD117 was positive in 47%, BCRP1 in 50% and CD133 in 5% of the tumours. CD117 showed diffuse membranous and/or cytoplasmic staining (fig. 1b, f and j), while BCRP1 displayed additionally nuclear staining and predominantly focal expression (fig. 1d, h and l). The alveolar epithelium of patients was negative for adult stem cell antigens (fig. 1i–l), whereas the tracheobronchial epithelium showed cytoplasmic expression of CD117 in the basal cells and in some ciliated cells and nuclear expression of BCRP1 in the majority of ciliated cells (fig. 1e–h).
Prognostic significance of SSEA-4
Within the median follow-up time of 91.4 months for SSEA-4-positive and 70.2 months for SSEA-4-negative patients, significantly more individuals with SSEA-4-positive than SSEA-4-negative tumours experienced relapse (79% versus 21%, p = 0.0005). All but one relapse were distant relapses (table 1). The 5-yr survival rate of SSEA-4-positive patients was 35% and that of SSEA-4-negative patients 80%. Cox proportional regression analysis revealed a 6.0-fold increased risk of relapse (hazard ratio (HR) 6.0, 95% CI 2.1–17.4; p = 0.001) and a 4.2-fold increased risk of disease-related mortality (HR 4.2, 95% CI 1.3–13.7; p = 0.017) in patients with SSEA-4-positive tumours (figs 5a and 5b). These patients showed a median disease-free survival of 11.1 months and median overall survival of 45.0 months, while the median disease-free and overall survival of patients with SSEA-4-negative tumours were not reached.
DISCUSSION
Basaloid lung cancer represents a rare variant of NSCLC and showed an unfavourable prognosis in a number of studies [1, 5, 6]. Brambilla et al. [5], who were the first to describe this variant, speculated as early as 1992 that basaloid NSCLC might derive from a pluri- or multipotent pulmonary stem cell. Although a few studies indicated the expression and unfavourable prognostic significance of embryonic and adult stem cell antigens, such as SSEA-1, Oct-4, CD133 and BCRP1 in NSCLC, analyses particularly in basaloid lung cancer are lacking so far [11, 12, 16, 17, 26, 27]. In our study analysing 38 patients with early-stage basaloid NSCLC, we found expression of the early ESC antigen SSEA-4 in 37% of the specimens and an association with a 6.0-fold increased risk of relapse and a 4.2-fold increased risk of disease-related mortality. While patients with SSEA-4-negative tumours showed a relapse rate of 21% and a 5-yr survival rate of 80%, patients with SSEA-4-positive tumours were recurrent in 79% of the cases and showed a 5-yr survival rate of 35%. Without stratification for SSEA-4, the relapse rate (42%) and 5-yr survival rate (62%) of the entire collective was similar to that reported for overall early-stage NSCLC or poorly differentiated SCC (PDSC) [1–6]. This is in contrast to the results of Brambilla and co-workers [1, 5, 6], who reported a significantly worse outcome of basaloid carcinomas as compared with other NSCLC or PDSC: While the latter displayed a 5-yr survival rate of 44% and 47–55% in stage I/II disease, respectively, the 5-yr survival rate of early-stage basaloid NSCLC was 10–27% [1, 5, 6]. Conversely, previous studies of Kim et al. [2] and Wang et al. [28] in two Asian collectives likewise demonstrated comparable relapse rates (33–55%) and 5-yr survival rate (50–57%) for basaloid carcinoma and PDSC. The results of our study might provide one of the reasons for the inconsistent findings of previous studies, as a random accumulation of patients with SSEA-4-positive tumours might result in worse prognosis of the entire study population with basaloid NSCLC. However, several other factors have to be considered for these discrepancies between the studies including technical aspects, investigator's variability in the identification of basaloid NSCLC, changes of the staging system, differences in the surgical procedure and post-operative treatment phase, differences in the interval and type of clinical assessments, as well as differences of the number of analysed subjects, the follow-up time and clinico-pathological characteristics of the study population. Moreover, the better treatment options for relapsed patients since 2000 might have levelled discrete survival differences in the more recent studies of Kim et al. [2], Wang et al. [28] and Waechter et al. [29].
Besides clinical significance, the expression of the early ESC antigen SSEA-4 in basaloid carcinoma and in basal cells of the tracheobronchial epithelium represents an important pathophysiological aspect and supports the stem cell hypothesis of Brambilla et al. [5]. In line with rapid downregulation of this glycolipid antigen in the embryonic carcinoma cell lines 2102Ep, BG01V, NTERA-2 and in blastocyst-derived ESC upon differentiation, we found no expression of SSEA-4 in well-differentiated tumour components of basaloid NSCLC [30–34]. Moreover, a recently established basaloid NSCLC cell line that shows properties related to stem cells, i.e. high tumorigenicity in the animal model and reassembling of the original tumour histomorphology, also displayed expression of SSEA-4 [25]. However, as defined model systems and conditions for the assessment of multi-lineage differentiation capacity of pulmonary cells are lacking so far, the nature of SSEA-4-positive cells in the tracheobronchial epithelium and in some basaloid NSCLC remains to be elucidated. As for the current study, the results suggest that a part of the basaloid carcinomas retain the embryonic antigen expression of their putative cell of origin, while others lack either any stem-cell antigens or show exclusively expression of adult stem cell markers. It may be speculated whether these carcinomas derive from more committed (stem) cells or undergo differentiation during tumour growth. The lack of expression of CD133 and BCRP1 in the basal cells of the tracheobronchial epithelium, but focal expression of BCRP1 in some SSEA-4-positive tumours suggests the capability of intratumoral differentiation.
As the representativeness of the TMA plays a critical role particularly in focally expressed antigens, such as BCRP1, the TMA was prepared according to recent validation data in lung cancer and mesothelioma that demonstrated that three to four tissue cores per sample are sufficient to produce reliable results [35–37]. Moreover, in half of the cases, the results of the TMA were compared with corresponding whole tissue sections and showed full concordance.
With expression of the early ESC antigen SSEA-4 in basaloid carcinoma and in the basal cells of the tracheobronchial epithelium, the transcription factor and proto-oncogene c-myc has to be discussed as a potential molecular driver of this NSCLC variant. c-myc represents one of the factors that is crucially involved in the early steps of somatic cell reprogramming and is the centre of a regulatory network that induces an embryonic gene expression profile, albeit with no pluripotency in cancer cells [38, 39]. Depending on the method, 5.6–80% of all NSCLCs show an amplification of this gene and 48–58% an overexpression [40–42]. Remarkably, in normal tracheobronchial epithelium, the highest expression values were reported for basal cells, and tumours arising from these natural progenitor cells might per se display ESC features [42]. To date, no study or subgroup analysis of basaloid lung cancer is known to the authors that would have analysed the status of c-myc or the genetic profile of this variant. Our own comparative genomic hybridisation and multiplex fluorescence in situ hybridisation analysis of two primary tumours and one lymph node metastasis of basaloid NSCLC that was conducted in the context of cell line establishment demonstrated a consistent amplification of 8q24, the gene locus of c-myc [25].
Despite this, the presence of tumour cells with ESC characteristics might provide novel therapeutic options and targets for NSCLC: glycolipid antigens, such as SSEA-4, are highly immunogenic and might be suitable for antibody-based therapy approaches, whereas embryonic signalling pathways, such as Wnt/β-catenin, hedgehog or notch, could be a target for small molecule inhibitors [43, 44].
Conclusions
This study demonstrated expression of the ESC antigen SSEA-4 in a fraction of basaloid NSCLC and in the basal cells of the tracheobronchial epithelium. Patients with SSEA-4-positive tumours showed a significantly increased risk of relapse and disease-related mortality. With respect to the limited sample size of this study, which precludes a definitive statement, SSEA-4 might represent a promising candidate for the identification of patients with basaloid NSCLC who will benefit from an adjuvant therapy after surgery.
Acknowledgments
The authors thank B. Walter (Tissue Bank, National Center for Tumour Diseases, Heidelberg, Germany) for excellent technical assistance.
Footnotes
Statement of Interest
None declared.
- Received December 23, 2011.
- Accepted May 30, 2012.
- ©ERS 2013