Biomarkers Being Identified for Improved Diagnosis of Nasopharyngeal Carcinoma Via Liquid Biopsy

Dhruvajyoti Roy, Ph.D

Dhruvajyoti Roy, Ph.D

Nasopharyngeal carcinoma (NPC) is predominantly diagnosed using invasive testing methods like endoscopic biopsy or tissue biopsy. However, new research presented at the 2019 ASCO Annual Meeting demonstrated that biomarkers have been identified for the early detection and monitoring of NPC using liquid biopsy.

In a phase I study, researchers obtained 168 blood samples to test cancer-specific DNA methylation patterns in cell-free DNA (cfDNA) for NPC methylation markers. Of the 168, 59 samples were from subjects who had already been diagnosed with NPC, and 57 out of 59 were found to be accurate diagnoses with an overall sensitivity of 97% using cfDNA. Researchers also collected 43 samples from healthy subjects and 14 from subjects with benign nasopharyngeal disease.

The results showed subjects with stage I to stage IV NPC had a sensitivity range from 92% to 100%. Those diagnosed with other cancers showed a total specificity of 86%, which amounted to 93% in both breast cancer and liver cancer samples. Results from the healthy and benign sample showed a combined specificity of 100%. From these data, the researchers were able to conclude that liquid biopsy was sensitive and specific for early detection of NPC. The researchers predict that the use of these biomarkers will improve outcomes for patients with NPC.¹

During an interview withTargeted Oncology, lead study author Dhruvajyoti Roy, PhD, the director of technology at Laboratory for Advanced Medicine, Inc., shared information about the study, commented on other liquid biopsy research presented by peers at ASCO and provided insight as to what these data mean for the future of cell-free DNA methylation.

TARGETED ONCOLOGY:This week you presented on the recent positive findings for detecting new types of cancer using liquid biopsy. Can you discuss the rationale for this study first?

Roy:[At ASCO], we presented the Nasopharyngeal Carcinoma (NPC) data. Although, [nasopharyngeal carcinoma] is fairly rare in the United States, but it is the 3^rdmost prevalent malignancy among menin South East China. Let me [first explain] what nasopharyngeal carcinoma is. It is a non-lymphomatous squamous cell carcinoma which is a distinct form of head and neck cancer. And it occurs in the epithelial lining of the nasopharynx.

As I mentioned, NPC is most prevalent in Southern China and it is [actually] the 4th most common cancer in Hong Kong, constituting 1 of the most prevalent malignancies among populations native to Southeast Asia, the Mediterranean Basin and the Arctic. Detection of NPC is hard and particularly the examination of nasopharynx is really challenging. Traditionally, people use an invasive nasal endoscopy which is not practical for early detection, screening, or even for the post-therapy disease monitoring.

So, that's why we believe that there's an unmet need and non-invasive tests from liquid biopsies have great potential to improve the overall survival rate. LAM has been analyzing methylation for years, and by analyzing methylation, which is an epigenetic process and is predictable in cancer and healthy patients, we can detect NPC early. We have already established several biomarkers for different cancer types, so we thought it would be ideal to develop DNA methylation-based biomarker for NPC. That's why we started to develop those biomarkers.

TARGETED ONCOLOGY: What were the findings you presented at the meeting?

Roy:So, [for] this particular study, we used almost 168 samples and among those, 59 subjects [were] diagnosed with nasopharyngeal carcinoma (stage I to stage IV). We also tested 14 subjects diagnosed with benign nasopharyngeal disease and 43 healthy donor samples to check specificity. In addition, we have also tested around 52 other cancer types to check the analytical specificity of those markers. What we’ve found so far with this study is that we had an overall sensitivity of 97%, with little difference between the sensitivity of detecting Stage I to Stage IV NPC (range 92% to 100%). For subjects diagnosed with other cancers, a total calculated analytical specificity was 86%. All healthy donor and benign nasopharyngeal disease samples were correctly identified as negative for nasopharyngeal carcinoma for a combined specificity of 100%. We demonstrated that the markers are very sensitive and specific for the detection of nasopharyngeal carcinoma. 

TARGETED ONCOLOGY: What is the significance of these findings?

Roy:There are several tests for NPC that are commonly used such as Nasoendoscopy and Epstein Barr virus (EBV) in the DNA of the cells. However, due to infection of virus, EBV is not a specific marker for NPC. Not everyone who has EBV will get nasopharyngeal cancer. Particularly, in the United States, people who have had an EBV infection never have long-term issues [associated] with that. So, we believe having our test will help for the early detection and monitoring of the disease. 

Due to the minimal and the non-specific local symptoms and the inaccessibility of the nasopharynx for routine examination, we believe that applying this liquid biopsy-based test will greatly increase the chances of effective treatment and appropriate monitoring of the disease. The subjects whodo not have tissue available or cannot undergo a new invasive biopsy after surgery,—these people can also [benefit] from this test. Our test provides the hope to detect nasopharyngeal cancer at an early stage when the treatment options available to the physician are more manifold. The patients’ health outcomes can be improved at lower cost.

TARGETED ONCOLOGY: Are there any next steps planned for this research?

Roy:These are our phase I data. So, we are planning to have additional validation of a large cohort NPC population with the objective of making an accurate test commercially available to the at risk population. Particularly, we would like to validate this assay in our CLIA/CAP-approved laboratory, before we can make it commercial with potential utility in clinical diagnostic and monitoring of treatment response. 

TARGETED ONCOLOGY: Liquid biopsy was a hot topic at ASCO this year. Could you discuss some of the other research you saw in this field?

Roy:I was talking with other oncologists and physicians during the conference and they agree that liquid biopsy is a very upcoming field and people would like to get more information related to liquid biopsy-based assays because there's a great potential for screening of the disease. It has potential for companion diagnostics as well. People can easily use these tests for monitoring treatment and drug responses. I also saw a couple of posters [at ASCO] related to detection and quantification of minimal residual disease (MRD) after the surgery and reoccurrence that employed liquid biopsy in the management of patients.

It has great potential, but some liquid biopsy based assays show challenges for early detection of cancer. Most of the companies you will see in the liquid biopsy field are focusing on sequencing somatic mutations using cell-free DNA, but the sensitivity of these methods will be low among patients with early-stage cancer because of limited number of recurrent. The sensitivity has remained a key obstacle in the development of these mutation-based methods for early detection. In this context, DNA methylation-based approach could be superior as a cancer specific marker to DNA mutation because of several frequently aberrant DNA methylation in gene regulatory regions which provide higher sensitivity .  

What I've also noticed at the conference that the early detection paradigm is shifting more toward DNA methylation. Researchers are working more on DNA methylation-based assays. I saw so many posters related to several cancer types using DNA methylation, including analysis of whole genome or specific methylation. The good thing about our technology is that while people are looking at the whole-genome DNA methylation, we have already identified and validated a few cancer-specific markers for major cancer types. We went through several databases including The Cancer Genome Atlas (TCGA) and implemented machine-learning algorithms/AI up front to find out and narrow down those hypermethylated markers so those could be useful for early detection.

After observing several research works, I realized that we are moving in the right direction and we got very good feedback from the top oncologists around the world. They agree that it is still relatively early days, but it has great potential, particularly in the field of early detection and treatment monitoring.

I saw several posters highlighted mutation-based panels to monitor drug responses. Researchers presented comprehensive genomic profiling by using multiple gene panels from different companies, whether [it is] 500 gene panels or more targeted gene panels like 30-75 genes. Those are not much useful for early detection but  helpful for the pharmaceutical companies to see how their drugs are working and also for therapy selection and monitoring. 

Other than that, for early detection researchers are also relying on CTCs, and EVs/exosomes from various body fluids. Despite different approaches developed to detect CTCs and exosomes, none of them completely meets the application requirements as loss of CTCs, low purity, and a narrow detection spectrum still need to be addressed. A few posters showed the potential of 2 exosomal cargoes. One is called protein cargo and the other is nucleic acid (RNA/DNA) cargo. Some studies have also been performed for evaluating the role of exosomal microRNAs (miRNAs) for cancer diagnostics because those are also other components of liquid biopsies. And other than that, a few posters also showed miRNAs profiling which added a new dimension to this already complex field of research, but from my experience, it's a bit hard to reach some kind of endpoint for clinical applications. You need lots of algorithms and AI to understand how these data can be useful. However, a significant amount of progress has been made towards advances in liquid biopsies by combining NGS and machine learning/AI and we will see a great expansion in its use in the next few years. 

Reference:

Dhruvajyoti Roy, David J Taggart, Lianghong Zheng, et al. Cell-free DNA methylation markers for noninvasive early detection of nasopharyngeal carcinoma.J Clin Oncol.2019;37(suppl; abstr e14537). http://abstracts.asco.org/239/AbstView_239_260949.html.