Expression of hypoxia-inducible factor-1 alpha and its prognostic significance in small-sized adenocarcinomas of the lung

Recent advancement in radiographic techniques, such as high-resolution computed tomography, has enabled us to detect small-sized adenocarcinomas of the lung [1,2]. Early detection may eventually lead to a decrease in disease mortality. However, the prognosis of patients with lung carcinoma is generally poor, even in patients with small-sized adenocarcinomas [3]. Noguchi et al. have conducted research to elucidate the histopathologic prognostic factors for small-sized adenocarcinomas of the lung. They showed that lung adenocarcinomas with active fibroblastic proliferation are more likely to metastasize and have a less favorable prognosis [4]. However, the molecular mechanisms underlying the aggressive behavior of small-sized adenocarcinomas of the lung are poorly understood. To improve the prognosis of patients with small-sized adenocarcinoma of the lung, new therapeutic strategies based on a better understanding of tumor behavior are needed. The transcription factor hypoxia-inducible factor-1 (HIF-1) consists of HIF-1 alpha and HIF-1 beta subunits. HIF-1 alpha reflects HIF-1 activity, because it is expressed rapidly in response to hypoxic conditions, while HIF-1 beta is expressed constitutively [5–7]. HIF-1 alpha protein has a very short half-life under normoxic conditions due to its continuous ubiquitination and proteosome-mediated degradation. Hypoxia stabilizes HIF-1 alpha protein by inhibiting ubiquitination and degradation [8]. Previous studies described that HIF-1 alpha plays a critical role in angiogenesis [9–11]. Therefore, we hypothesized that the expression of HIF-1 alpha may have prognostic value in patients with small-sized adenocarcinomas of the lung. In this study, we investigated the prognostic significance of HIF-1 alpha and its relationship with clinicopathological factors and angiogenesis factors, such as vascular endothelial growth factor-A (VEGF-A) and VEGF-C. Tumor specimens were obtained from 78 patients who had small-sized adenocarcinoma of the lung (maximum tumor dimension ≤ 2 cm). All patients were treated by standard surgical therapy alone at the Fukuoka University School of Medicine (Fukuoka, Japan) between 1993 and 2002. Informed consent was obtained from each patient. Disease stage was determined based on the tumor, node, metastasis classification of the International Union Against Cancer [12]. All patients underwent contrast-enhance computed tomography scans within 1 month prior to surgery. Histological typing was determined according to the World Health Organization classification [13]. The smoking index was defined as the product of the number of cigarettes smoked per day and the number of years of smoking. We defined cN2 as mediastinal lymph node(s) greater than 1.0 cm in shortest dimension, as determined by preoperative conventional computed tomography. Cases were classified as either localized bronchioloalveolar carcinoma (LBAC) or conventional adenocarcinoma according to the major morphologic pattern. The LBAC group consisted of patients with bronchioloalveolar growth pattern tumors that had no evidence of vascular, pleural, or stromal invasion. The conventional adenocarcinoma group consisted of patients with invasive adenocarcinoma, encompassing mixed subtype, acinar, papillary, and solid adenocarcinoma. Lymphatic invasion and vascular invasion indicated tumor cells identifiable in the lymphatic and vascular vessel lumen, respectively. Scar grades were classified as follows: grade 1, tumor had foci of alveolar collapse with resulting condensation of elastic fibers but no or minimal fibroblastic tissue with collagen; grade 2, tumor had fibroblastic tissue with a small amount of collagen fibers; grade 3, tumor had fibroblastic tissue with moderate or abundant amount of collagen fibers; and grade 4, tumor showed hyalinization [14]. The following clinicopathologic factors were analyzed to determine if they impacted patient survival: age, gender, smoking index (<400 vs ≥400), serum carcinoembryonic antigen (CEA) level (<5.0 mg/ml vs ≥5.0 mg/ml), clinical T factor (cT) (cT2-4 vs cT1), clinical lymph node status (cN) (mediastinal node involvement as cN2 vs less extensive as cN0-1), histological type of tumor (LBAC vs conventional), pleural involvement (p0-1 vs p2-3), lymphatic invasion (positive vs negative), vascular invasion (positive vs negative), pathologic N status (pN: pN2-3 vs pN0-1), and degree of fibrotic scarring (scar grade 1–2 vs grade 3–4). Formalin-fixed and paraffin-embedded tissues were deparaffinized. Tissues used for immunohistochemical staining were subsequently microwaved in citrate buffer. Tissues were preincubated with 0.3% hydrogen peroxide and protein blocking solution. For HIF-1 alpha (H-206), a 1:200 dilution of rabbit polyclonal antibody (Santa Cruz Biotechnology, Santa Cruz, CA, USA) was added to the sections. For VEGF-A (clone JH121), a 1:100 dilution of mouse monoclonal antibody (Uptake, New York, NY, USA) was added to the sections. For VEGF-C, a 1:200 dilution of rabbit polyclonal antibody (Zymed Laboratories Inc., South San Francisco, CA, USA) was added to the sections, and they were incubated overnight at 4 °C. After incubation with secondary antibodies, the streptavidin–biotin–peroxidase complex was added, and peroxidase activity was visualized with 3-amino-9-ethylcarbazole. The tumors were scored by three investigators who were blinded to the patients’ clinical data. For HIF1-alpha, VEGF-A, and VEGF-C, microscopic examinations were scored as follows: weak, <10% of cells staining; and strong, ≥10% of cells staining. These evaluation criteria are described in previous reports [15–17]. The survival rates were calculated by the Kaplan–Meier method [18], and univariate analyses were performed with the log-rank test [19]. StatView Version 5.0 (SAS Institute Inc., Cary, NC, USA) was used to perform statistical analyses. p-values less than 0.05 were considered statistically significant. HIF-1 alpha expression was examined in a series of small-sized adenocarcinomas (maximum dimension ≤ 2 cm) using a polyclonal antibody against a recombinant protein corresponding to amino acids 575–780, which map near the carboxy terminus of human HIF-1 alpha. HIF-1 alpha was observed almost exclusively in the cytoplasm of lung tumor cells. Little or no immunoreactivity for HIF-1 alpha was found in normal alveolus and LBAC components (Fig. 1A) of almost all cases. Immunoreactive HIF-1 alpha was frequently found in the solid components of adenocarcinoma with mixed subtype (Fig. 1B). Localization of hypoxia-inducible factor-1 alpha (HIF-1 alpha) in lung adenocarcinoma. (A) In bronchioloalveolar carcinomas, staining for HIF-1 alpha was absent or minimal. (B) In adenocarcinomas with mixed subtypes, HIF-1 alpha was preferentially found in the cytoplasm of cancer cells and in the solid component of the cancer stroma (magnification ×200). Patient clinicopathologic characteristics are shown in Table 1. Relationships between HIF-1 alpha expression and clinicopathologic characteristics are shown in Table 2. Strong expression of HIF-1 alpha was significantly higher in cases with vascular invasion, lymph node involvement, and VEGF-A expression. Patient characteristics (n = 78) Relationship between HIF-1 alpha expression and clinicopathological characteristics (n = 78) The Cox proportional hazards model analysis for overall survival revealed that clinical N status (p = 0.0057), pathological N status (p = 0.0003) and strong expression of HIF-1 alpha (p = 0.0036) were associated with reduced survival (Table 3). Kaplan–Meier survival curves according to HIF-1 alpha expression are shown in Fig. 2. Patients with strong HIF-1 alpha expression had a significantly lower survival curve (p < 0.05). Univariate analysis of survival in lung adenocarcinoma of small size We observed strong HIF-1 alpha expression in 23 (29%) of 78 cases. HIF-1 alpha was preferentially expressed in the solid components of conventional adenocarcinomas and absent or minimal in all LBAC cases. These observations may indicate that LBAC is under normoxic conditions because HIF-1 alpha expression correlated with hypoxia in a previous study [20–22]. On the other hand, conventional adenocarcinoma is under hypoxic conditions. Hypoxia may play an important role in the adenocarcinoma development sequence in which atypical adenomatous hyperplasias progresses to LBAC with central fibrotic foci, so-called sclerosing BAC [4]. To our knowledge, this is the first study indicating that HIF-1 alpha is a potentially important marker for small-sized adenocarcinomas of the lung. We should try to analyze prognostic power of HIF-1 alpha expression on a multivariate basis in a large cohort of convert population. Studies toward this aim are now under way. A previous report indicated that HIF-1 alpha expression in non-small cell lung cancer was marginally associated with T-stage and poor prognosis [23]. In the present study, we investigated the associations between the strong expression of the HIF-1 alpha and clinicopathologic characteristics. We found that strong expression of HIF-1 alpha was associated with vascular invasion, pathologic stage, and VEGF-A expression. These findings suggest the significance of HIF-1 alpha as a biomarker of tumor angiogenic potential for small-sized adenocarcinomas of the lung. This study revealed that HIF-1 alpha expression is more sensitive prognostic marker than VEGF-A expression in small-sized lung adenocarcinomas. This result may be explained by the fact that HIF-1 alpha is an upstream signal of VEGF-A in the angiogenesis transduction pathway. The angiogenetic cascade is that HIF-1 alpha protein is transported into the nucleus, where it forms a heterodimer with HIF-1 beta. The HIF-1 heterodimer then binds to its DNA-recognition sequence, 5′-RCGTG-3′, which is present within the hypoxia response element of target genes, resulting in the activation of the gene encoding VEGF-A [9–11]. It may be argued that strong HIF-1 alpha expression is associated with tumor angiogenesis, HIF-1 alpha could be an important target for preventing tumor angiogenesis. Remarkable progress has been made in molecular target therapy of non-small cell lung cancer, such as the development of epidermal growth factor receptor (EGFR) inhibitors, in the past few years [24]. It is now accepted that the EGFR signal transduction network plays an important role in multiple tumorigenic processes including angiogenesis. Luwor et al. [25] described that the antiangiogenic mechanisms of cetuximab (C225), an EGFR inhibitor, include its ability to inhibit HIF-1 alpha expression via the PI3-K/Akt signaling pathway and to decrease HIF-1 alpha protein synthesis. Therefore, cetuximab might be useful for an adjuvant therapy of small-sized lung adenocarcinomas that express HIF-1 alpha. Thus, it would be interesting to explore whether a combination of surgery and cetuximab may also lead to an efficient strategy for the treatment of lung adenocarcinomas with solid component, especially the component under hypoxic condition. Immunohistochemical staining of HIF-1 alpha, along with examination of metastatic potential via vascular pathways, may be valid defining a subpopulation of patients with small-sized adenocarcinoma of the lung whose tumors have aggressive angiogenesis potential, such as vascular invasion, lymph node involvement, and vascular endothelial growth factor-A expression. The authors thank Takehito Kawakami, MD and Shinichi Maekawa, MD for their valuable technical assistance.