Ylstra received a bachelors in Physics in 1988 from the University of Amsterdam (UvA, NL) and commenced with a masters in Genetics at the Vrije Universiteit (VU, NL).

In 1995 he received his PhD degree from the same university for work performed at the Universities of Vienna (A), and Nijmegen (NL), Plant Research International (NL) and The John Innes Institute (UK). He was awarded a 2-year NATO fellowship for his post-doctoral work at the University of California in Berkeley (USA) on genome-wide RNA screens.

From 1998 he worked at the Comprehensive Cancer Center of the University of California, San Francisco (USA) and was involved in the development of DNA arrays for copy number analysis.

In 2001 he accepted his current position at the Amsterdam UMC, Cancer Center Amsterdam heading the Tumor Genome Analysis Core (tgac.nl). The core has a strong genome laboratory and bioinformatics (pre-)processing track record, with an emphasis on Next Generation Sequencing (NGS) DNA analysis from archival formalin fixed and parrafin embedded (FFPE) materials and liquid biopsies.

ABSTRACT

Title: A new biomarker for bevacizumab effectiveness in mCRC – hope or hype?

Summary: Bevacizumab (trademark Avastin) is used in the treatment of patients with metastatic ovarian, cervical, breast, lung, kidney and colorectal cancer with a median survival gain of up to approximately 4 months. Although side effects are generally limited and manageable, they can be serious or even fatal.

In addition, bevacizumab is an expensive drug that costs between € 20,000 and € 40,000 per patient per year. Unnecessary treatment therefore entails unnecessary costs, while cost-effectiveness is questionable. Consequently, bevacizumab is not reimbursed for metastatic colorectal cancer(mCRC) in the UK and is not prescribed in the Netherlands in 50-60% of cases, despite national and international guidelines. The urgency for a biomarker is therefore high, which is recognized by the European Association of Medical Oncology (ESMO).

Since it’s approval, numerous candidate biomarkers, have been proposed but were based on small exploratory studies and validation in larger independent patient series failed.

In 2014, we presented the results of a genome-wide screening of 119 well-annotated mCRC tumors from patients treated with bevacizumab in the Dutch CAIRO2 phase III clinical study. Loss of (parts of) chromosome 5q, 17q and 18q correlated significantly with long-term progression-free survival (PFS), while this correlation was not seen for 186 patients of the CAIRO phase III clinical study of patients did not receive bevacizumab (Nat Commun. 2014 Nov 14;5:5457).

In 2018 presented the validation of the 18q marker on the basis of 600 archived tumor samples from patients with metastatic colorectal cancer from Ireland, Italy, Netherlands and Belgium, and demonstrated that a third of mCRC patients would benefit less from treatment with bevacizumab (J Clin Oncol. 2018 Jul 10;36(20):2052-2060). Loss of chromosome 18q was most commonly associated with long-term PFS and OS (PFS difference, 49 days without vs. with bevacizumab).

Alternatively, we showed that chromosomal instability, which has been linked to chromosome 18q loss (Nature. 2013 Feb 28;494(7438):492-496), could serve as a marker for bevacizumab efficacy (Nat Commun. 2018 Oct 5;9(1):4112).

Hope or hype? Historically, less than 1% of published biomarkers make it to clinical practice. What evidence is required for oncologist and patient alike to withhold a standard treatment modality, in this case bevacizumab? Is an analysis with retrospective samples allowed? What trial options are available; non-inferiority, superiority, interaction analysis?

We propose a retrospective analysis of a trial of patients treated with vs. without bevacizumab and like to discuss if significant results would be sufficient to convince clinician and patient to abstain from bevacizumab complementation of their treatment.