Virtual Library

Start Your Search

Sjoerd H. Van Der Burg



Author of

  • +

    P2.04 - Immuno-oncology (ID 167)

    • Event: WCLC 2019
    • Type: Poster Viewing in the Exhibit Hall
    • Track: Immuno-oncology
    • Presentations: 2
    • Moderators:
    • Coordinates: 9/09/2019, 10:15 - 18:15, Exhibit Hall
    • +

      P2.04-06 - Blood-Based Multiplex Kinase Activity Profiling as a Predictive Marker for Clinical Response to Checkpoint Blockade in Advanced NSCLC (ID 1322)

      10:15 - 18:15  |  Author(s): Sjoerd H. Van Der Burg

      • Abstract

      Background

      Only a minority of non-small-cell lung cancer (NSCLC) patients benefit from treatment with immune checkpoint inhibitors (ICIs), therefore, there is an urgent need for response prediction. Previously, the potential of using tyrosine kinase activity profiling of baseline peripheral blood mononuclear cells (PBMC) was demonstrated in an analysis with ICI treated advanced melanoma patients [1]. Here, we apply this methodology to evaluate the predictive value for response to ICIs in NSCLC.

      Method

      59 ICI naïve advanced NSCLC patients treated with PD-1 blockade were included in this exploratory analysis of the prospective immuno-monitoring study (MULTOMAB; NTR7015). PBMC were isolated from patient blood samples obtained prior to treatment with ICIs. Kinase phosphorylation signatures of PBMC lysates were measured using a micro-array, comprising of 144 different peptides derived from sites that are substrates for protein tyrosine kinases. Correlation of the profiles with progression free survival (PFS) and overall survival (OS) was analyzed using univariate cox-regression. Predictive multivariate (GLMnet) analysis was performed by binary survival grouping of patients with a cut-off at 140 days for PFS and 365 days for OS. The predictive performance of the models was estimated using cross validation. Multiplex flow cytometry, enumerating 18 immune cell subsets and assessing expression for 28 T cell markers, was performed for a selection of patients to gain additional insight in the immune profile [2].

      Result

      Univariate Cox regression showed significant correlation of phosphorylation signal with PFS for 7 peptides (p < 0.01, False Discovery Rate [FDR] = 10%), and with OS for 34 peptides (p < 0.01, FDR = 2%). Evaluation of the predictive value of GLMnet models resulted in estimates for the Correct Classification rate (CCR) of 67-70% for PFS and 67-73% for OS. When the cross validated predictions of the models were used to categorize the patients in a predicted-high-risk and a predicted-low-risk group, this resulted in a significant difference in survival between these groups. The predicted-low-risk group displayed significant better median PFS and OS than the predicted-high-risk group ([246 vs. 56 days; HR 2.3, 95%CI 1.2-4.7, p=0.02] and [488 vs. 171 days; HR 2.7, 95%CI 1.1-6.6, p=0.03], respectively).

      Conclusion

      Similar to melanoma, kinase activity profiles of baseline PBMC samples of advanced NSCLC patients can predict the response to PD-1 blockade. This assay may serve as a predictive biomarker for response to anti-PD-1 therapy. Involvement of immune receptor kinases is being investigated. An independent validation study is underway.

    • +

      P2.04-47 - Tumor Mutational Load, CD8+ T Cells, Expression of PD-L1 and HLA Class I to Guide Immunotherapy Decisions (ID 1259)

      10:15 - 18:15  |  Author(s): Sjoerd H. Van Der Burg

      • Abstract

      Background

      A minority of NSCLC patients benefit from anti-PD1 immune checkpoint inhibitor therapy. A rational combination of biomarkers is needed. The value of using a series of mechanism-of-action based parameters was studied for response prediction of immunotherapy: tumor mutational load (TML), CD8+ T cell infiltration, HLA class I expression and the currently used PD-L1 tumor proportion score.

      Method

      Patients were prospectively included between April 2016 and August 2017, and retrospectively analyzed. Metastatic NSCLC patients (n=30) with sufficient archival tissue, obtained prior to the first nivolumab administration, were selected. Response was assessed by RECIST v1.1. Progression-free survival (PFS) and overall survival (OS) were analyzed by Kaplan-Meier methodology. TML was determined using a next-genome sequencing panel (409 cancer-related genes). Immunohistochemistry was performed to score PD-L1, total CD8+ T cell infiltration and HLA class I.

      Result

      In 30 patients with adenocarcinoma (67%) or squamous cell carcinoma (33%), high TML was significantly associated with better PFS (p=0.004) and OS (p=0.025). Interaction analyses revealed that patients with both high TML and high total CD8+ T cell infiltrate (p=0.023) or no loss of HLA class I (p=0.026), patients with high total CD8+ T cell infiltrate and no loss of HLA class I (p=0.041) or patients with both high PD-L1 and high TML (p=0.003) or no loss of HLA class I (p=0.032) were significantly associated with better PFS. Unsupervised cluster analysis based on the four markers revealed three sub-clusters, of which cluster 1A was overrepresented by patients with progressive disease (15 out of 16), with significant effect on PFS (p=0.007).

      Conclusion

      This proof-of-concept study suggests that a combination of PD-L1 expression, TML, CD8+ T cell infiltration and HLA class I expression function as a better predictive biomarker for the response to anti-PD-1 immunotherapy and PFS. Consequently, refinement of this proposed set of biomarkers and validation in a larger set of patients is warranted.