Start Your Search
L. De Vogel
MINI 08 - Prognostic/Predictive Biomarkers (ID 106)
- Event: WCLC 2015
- Type: Mini Oral
- Track: Biology, Pathology, and Molecular Testing
- Presentations: 1
- Moderators:T.E. Stinchcombe, N. Pavlakis
- Coordinates: 9/07/2015, 16:45 - 18:15, Mile High Ballroom 4a-4f
MINI08.02 - Prediction of Response to Pemetrexed in NSCLC by Immunohistochemical Phenotyping Based on Gene Expression Profiles (ID 2793)
16:45 - 18:15 | Author(s): L. De Vogel
A major challenge in the treatment of advanced non-small cell lung cancer (NSCLC) is to identify specific tumor properties that predict response to chemotherapy. Although thymidylate synthase (TS) immunohistochemical (IHC) staining has been extensively studied as a predictive marker for pemetrexed (PEM) sensitivity, its clinical value remains limited. We investigated IHC stainings of different molecular markers linked to the folate metabolic pathway (FMP) identified with gene expression profiling (Hou et al, JTO 2012;7:105-114). We used a population with advanced NSCLC treated with PEM for external validation.
Resected tumor samples from PEM-naïve NSCLC patients were collected. Gene expression profiling with respect to predicted sensitivity to PEM was based on genes related to FMP. Based on differentially expressed genes, patients were divided into predicted responders (Rs) and non-responders (NRs). Genes showing a strong correlation with these FMP genes and for which IHC stainings were commercially available, were selected for measurement of corresponding protein expressions by IHC stainings. A semiquantitative scoring method was applied, which was used to construct a prediction model for response to PEM. Subsequently, a retrospective cohort of patients with advanced NSCLC was selected, who had received at least two cycles of PEM-based chemotherapy as first-line treatment. IHC staining scores for the same proteins were obtained from tumor tissue. The performance of the prediction model was tested in this population.
From 91 patients resected tumor samples were collected. The majority of patients had early or locally advanced NSCLC (96.3%). Gene expression profiling revealed five markers that showed mRNA levels strongly correlating to FMP genes mRNA levels: TPX2, CPA3, EZH2, MCM2 and TOPO2a. Of 63 patients IHC staining scores of these markers were obtained, which all correlated to their corresponding mRNA levels. The scores were significantly different between predicted NRs and Rs (p<0.05). Testing the IHC markers showed an optimized prediction model with CPA3 (OR=1.71, 95%CI (0.94-3.08)), EZH2 (OR=0.57, 95%CI (0.35-.093)) and TPX2 (OR=0.55, 95%CI (0.29-1.03)) included. With this model 86.5% of the predicted Rs and 72.7% of the predicted NRs were correctly classified. The ROC showed an AUC of 0.883 representing a good discriminatory performance. In the external study population (n=23) the majority of patients had metastatic NSCLC (95.7%). Partial response (PR) was established in 26.1%. Considering patients with PR as responders the prediction model classified 16.7% of the observed Rs and 88.2% of the observed NRs correctly. The ROC showed an AUC of 0.750.
Using external validation this prediction model with IHC staining of FMP correlated markers shows a good specificity, but lacks sensitivity. Again this study shows the limited value of IHC markers as response predictors for PEM in clinical practice. This may be ascribed to the poor relation between IHC and protein activity but the biological significance of FMP genes may also be less important than other factors influencing PEM activity, like pharmacodynamics of PEM e.g. the formation of metabolites. Metabolomics may offer better understanding in cellular processing of PEM and could provide new insights for tailored chemotherapy. Supported by an unrestricted grant from Eli-Lilly, the Netherlands
Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.