Project title: Emergence of high-grade and treatment-resistant pancreatic NET subclones

Description

Ziv will trace the evolution of pancreatic NETs to a higher grade and more aggressive forms using techniques used to study evolutionary biology.

What problem/questions will researchers try to answer?

Tumor grade remains the most important prognostic variable driving patient outcomes but little is known about how tumor grade changes over time and in response to treatment. Ziv will try to determine if higher grade tumors evolve from lower-grade tumors over time or arise independently and if certain treatments drive tumors to a higher grade.

Why is this important?

Understanding how aggressive tumors arise can inform one of the most pressing dilemmas in current pancreatic NET clinical management including treatment sequencing, the role of surgical debulking and locoregional therapy and surveillance.

What will researchers do?

Analyses of whole-exome and targeted deep sequencing of specimens from multiple time points will be performed to reconstruct tumor cell histories to identify and study why certain tumor cells become more aggressive.

How might this improve the treatment of NETs?

If higher grade tumors can be identified, these tumors can potentially be targeted and eradicated earlier in the course of the disease.

What is the next step?

Data generated by this study will provide the community with information about how and when aggressive tumor cells of pancreatic NETs arise, which can then be used to discover new biomarkers of disease aggressiveness and to test treatment sequencing strategies and inform new therapeutic strategies.

Outcomes:

Well-differentiated pancreatic neuroendocrine tumors (pNETs) are a group of cancers with highly variable clinical course. The most important prognostic variable is tumor grade. It has long been known that some pNETs can transform over time to higher grade and that some pNETs can have mixed grade tumors with regions of low and high grade in the same tumor. The relationship between high grade and low grade regions is unknown. Our preliminary data suggests that at least in some cases, high-grade subclones arise from very early progenitors and evolve in parallel with other low-grade subclones. It is unclear if this is a ubiquitous pattern or if high-grade tumors can also evolve later in the tumor history. It is unclear whether certain treatments can select for high-grade subclones. This research study aimed to resolve these issues by using an inference algorithm that can take genomic information from patients whose tumors were sampled over multiple time points and infer the tumor phylogeny. We wanted to determine how often high-grade tumors emerged, whether these high grade tumors were related phylogenetically to the original tumor and whether certain treatments were associated with grade change.

First, we found that transformation to high grade tumor was very common (over 75% were either transformed or in the process of transforming) and that the high-grade components were not clonally distinct. We also found that in some cases high-grade tumors were related to very early progenitors of the original tumor implying a highly branched pattern with early divergent evolution. Early high-grade emergence was associated with patients who had undergone surgical debulking, potentially providing an evolutionary perspective to explain the well-established but still poorly understood clinical phenomenon of improved outcomes in pNET patients who undergo tumor debulking. Some driver mutations were associated with high grade components but there was no ubiquitous pattern of genetic alteration associated with high grade.

Our study has established that pNETs transform commonly and that high-grade components are related to the original tumor rather than multi-clonal. This study underscores the importance of surveillance of pNETs. It also provides a mechanistic explanation for tumor debulking of pNETs. Given their inherent heterogeneity, the study also supports repeated sampling and further studies to use imaging and serum biomarkers to identify transformation early. We are now conducting single cell experiments to better delineate the subclonal architecture within samples.

« Back to all funded research projects