1033P - Hypoxia and metabolism gene expression profile predicts poor survival in head and neck carcinoma

Date 01 October 2012
Event ESMO Congress 2012
Session Poster presentation III
Topics Head and Neck Cancers
Translational Research
Presenter Sophie Gouerant
Authors S. Gouerant1, S. Mareschal2, J.M. Picquenot3, A. Berghian3, M. Cornic3, O. Choussy4, M. Benard5, F. Di Fiore6, F. Jardin2, F. Clatot2
  • 1Medical Oncology, Centre Henri Becquerel, Rouen/FR
  • 2Unité Inserm 918, Université de Rouen, Rouen/FR
  • 3Pathology, Centre Henri Becquerel, Rouen/FR
  • 4Head And Neck Surgery, CHU Charles Nicolle, Rouen/FR
  • 5Plateforme Primacen, Université de Rouen, Rouen/FR
  • 6Digestive Oncology Unit, Centre Henri Becquerel, Rouen/FR

Abstract

Background

Despite multimodal treatments combining surgery, radiotherapy, chemotherapy and now targeted therapies, head and neck squamous cell carcinomas (HNSCC) remain of poor prognosis. Hypoxia being an important biological characteristics of many HNSCC, we aimed to assess the prognostic value of the expression of 42 genes related to hypoxia and metabolism in non metastatic HNSCC.

Patients and methods

Expression of 42 genes was retrospectively analyzed by high throughput real-time PCR in 80 patients treated for a non metastatic HNSCC. Fresh tissue samples were collected at diagnosis. After extraction, RNA quality was evaluated by an Agilent 2100® Bioanalyzer. After reverse transcription of total RNA, cDNA were analysed using a 1536 LightCycler®. An unsupervised hierarchical clustering analysis was performed. Usual prognostic factors and clustering analysis results were related to overall survival (OS).

Results

Only 61 patients presented suitable quality RNA for gene expression analysis. With a median follow-up of 39.4 months, 41 patients were alive and 20 were dead of cancer evolution. Among usual prognostic factors, capsular rupture of involved nodes and lymphovascular invasion were related to poor OS (p = 0.008, p = 0.03, respectively). Gene expression profile distinguished 2 groups. The group 1 composed of 34 patients (7 dead, 27 alive) had a better OS than the group 2 composed of 27 patients (13 dead, 14 alive) (2 years OS = 85.3% and 2 years OS = 63.0%, respectively, p = 0.02). The group 1 had frequent overexpression of genes related to inflammation (SDF1, CXCR4, EP4). The group 2 had frequent overexpression of genes related to lactate metabolism (LDHA, GLUT1, HK2), hypoxia (HIF1, BNIP3) or pH regulation (CA9, MCT1). Multivariate analysis based on the usual prognostic factors and the clustering analysis confirmed that capsular rupture, lymphovascular invasion and gene expression profile were related to poor OS (p = 0.005, p = 0.02, p = 0.003, respectively).

Conclusion

In this cohort, clustering analysis underlined the prognostic value of hypoxia and lactate metabolism gene expression. If confirmed by further studies, tumor metabolism in HNSCC may become a promising target for anticancer drugs.

Disclosure

All authors have declared no conflicts of interest.