167O - Prognostic and predictive biomarkers for ACT (adjuvant chemotherapy) in resected non-small cell lung cancer (R-NSCLC): LACE-Bio

Date 27 September 2014
Event ESMO 2014
Session Biomarkers and tumour heterogeneity
Topics Anti-Cancer Agents & Biologic Therapy
Non-Small-Cell Lung Cancer, Early Stage
Pathology/Molecular Biology
Presenter Lesley Seymour
Citation Annals of Oncology (2014) 25 (suppl_4): iv58-iv84. 10.1093/annonc/mdu326
Authors L.K. Seymour1, G. Le Teuff2, M. Tsao3, E. Brambilla4, F.A. Shepherd5, J. Soria6, R. Kratzke7, S. Graziano8, J. Douillard9, R. Rosell10, A. Reiman1, B. Lacas11, A. Bourredjem2, T. Le Chevalier12, R. Pirker13, M. Filipits14, P. Hainaut4, P.A. Janne15, J. Pignon2
  • 1Queens University, NCIC Clinical Trials Group, Cancer Research Institute, K7L 3N6 - Kingston/CA
  • 2Service De Biostatistique Et D'epidemiologie, Institut de Cancerologie Gustave-Roussy, Villejuif/FR
  • 3Pathology Department, University Health Network, Princess Margaret Cancer Centre, Toronto/CA
  • 4Departement D’anatomie Et Cytopathologie, Centre Hospitalier Universitaire de Grenoble, Grenoble/FR
  • 5Department Of Medical Oncology, Princess Margaret Cancer Centre, University Health Network, M5G 2M9 - Toronto/CA
  • 6South-paris University, Departement de Médecine Gustave Roussy, 94805 - Villejuif CEDEX/FR
  • 7Oncology, University of Minnesota, Minneapolis/US
  • 8Hematology And Internal Medicine, SUNY Upstate Medical University Hospital, Syracuse/US
  • 9Medical Oncology, ICO Institut de Cancerologie de l'Ouest René Gauducheau, Nantes/FR
  • 10Medical Oncology Service, Catalan Institute of Oncology, Hospital Germans Trias i Pujol, 08916 - Badalona/ES
  • 11Service De Biostatistique Et D’épidémiologie, Institut Gustave-Roussy, Villejuif/FR
  • 12Dept. Of Medicine, Institut Gustave Roussy, FR-94805 - Villejuif CEDEX/FR
  • 13Department Of Medicine, Medical University of Vienna, AT-1090 - Vienna/AT
  • 14Medicine, Medical University of Vienna, Vienna/AT
  • 15Thoracic Oncology, Dana-Farber Cancer Institute, 02215 - Boston/US

Abstract

Aim

The precise selection of patients for ACT is critical but there remain no validated molecular tools which are prognostic for relapse or predictive of benefit from ACT.

Methods

LACE-Bio, based on the LACE (Lung Adjuvant Cisplatin Evaluation) meta-analysis project, includes a fully annotated database and tissue bank from 4 randomised trials comparing ACT to non-treated control (IALT(1), ANITA(2), NCIC CTG JBR-10 (3) and CALGB 9633(4). It contains ∼1500 samples, including frozen tissue (JBR.10). Histochemical/immunohistochemical (HC) biomarkers shown in one trial to have a significant prognostic and/or predictive effect on overall survival, were cross-validated in the 3 other trials; when only a trend (T) for such effects was observed we performed a pooled analysis combining all 4 trials. All statistical analyses were conducted by the meta-analysis unit at Gustave Roussy. Biomarker assays for validation/pooled analyses were in general conducted by the group reporting the original biomarker results

Results

Numerous issues were encountered during the attempted validation of promising assays, including method of fixation, storage, the use of TMAs vs. sections, stored vs. fresh sections, and reagent/antibody batches. Despite meticulous methodology, the majority of the promising biomarkers (other than mutations) could not be confirmed as of value.

Marker Trial Predictive? Prognostic? Confirmed?
ERCC1 IALT Yes Yes No
LYMPHOID INFILTRATE IALT No Yes Yes
MUCIN CALGB No Yes PROGNOSTIC (DFS)
ß-TUBULIN JBR10 Trend Yes PROGNOSTIC
P27 IALT Yes Yes No
FASL IALT Trend No No
FAS/FASL IALT Yes Yes No
BAX IALT Trend No No

Conclusions

Although inexpensive and widely available, HC based assays from single trials may be misleading and require validation before being implemented. Of particular concern are HC biomarkers which we could not validate, but which are being used as selection tools in open trials. LACE-Bio-2 is evaluating potential genomic biomarkers using next-generation sequencing that may allow more precise selection of NSCLC patients for ACT in NSCLC.

Disclosure

All authors have declared no conflicts of interest.