New blood-based diagnostic tools for lung cancer

Globally, lung cancer is the top contributor to cancer-related fatalities. The biggest challenge lies in the fact that symptoms manifest predominantly in the later stages of the disease, by which point it may be beyond treatment. Recently, hope has emerged from the promised impact of lung cancer screening to detect lung cancer earlier, and the remarkable advances in the fight against late-stage disease with new immunotherapy drugs.

Hummingbird is one of a handful of companies developing simple blood tests – liquid biopsies – that can diagnose diseases early or provide information for precision medicine. “The challenge in detecting cancer early has been high enough sensitivity,” says Timothy Rajakumar, vice president of medical sciences at Hummingbird Diagnostics, a biotechnology company based in Heidelberg, Germany. Low sensitivity means that cancers that are present may be missed. 

Hummingbird’s approach differs from most others in that it looks at both the disease and the body’s response to it “by merging the analysis of disease and immune system derived small RNAs,” explains Jochen Kohlhaas, CEO of Hummingbird.

“In contrast to tests that rely on identifying circulating tumor cells, cell-free DNA or RNA shed by tumors into the bloodstream, Hummingbird’s research suggests that “the combined analysis of both tumor- and immune derived signals allows us to achieve the required sensitivity that we need for a screening test,” explains Rajakumar. The intended impact of this test is evident as, “Early cancer detection remains the most effective strategy to reduce mortality associated with lung cancer,” says Kohlhaas.

New hope in advanced disease

The Hummingbird Team also realized that their unique liquid biopsy approach may have utility in late-stage disease. “Advanced-stage lung cancer is, unfortunately, a devastating condition. However, recently, new therapies have been developed, particularly immunotherapies which are beginning to make a difference,” says Rajakumar.

Despite this progress, existing tools to match precision treatments to patients need further development. “Only about 30% of patients with late-stage lung cancer will achieve a positive response to immunotherapy. And amongst the other 70%, some may even come to harm through hyper-progression. So it's really critical that we can predict which patient should get immunotherapy versus which patient may be better off with a different combination,” says Rajakumar.

A study by the National Cancer Institute found that deaths from non-small cell lung cancer have been gradually declining since personalized therapies became available (1). More significantly, researchers attribute the decline specifically to the advent of therapies that use genetic analysis of tumor tissue to help oncologists match cancers with appropriate treatments. 

For instance, patients with advanced lung cancer displaying elevated levels of programmed death-ligand 1 protein (PD-L1) in their tumor cells are more likely to benefit from immunotherapy drugs specifically designed to block this protein's function, empowering their immune systems to combat cancer.

However, only around 30 to 40% of patients identified through this assay respond to the treatment, which can also result in significant and, at times, life-threatening side effects.

To address this shortfall, Hummingbird is using their novel liquid biopsy strategy, that incorporates information from the immune system, to develop a complementary diagnostic tool for advanced-stage non-small cell lung cancer (NSCLC), a subtype accounting for 75 to 85% of all lung cancers. 

Their innovative diagnostic revolves around measuring levels of specific microRNAs (miRNAs) – a subclass of small RNAs (approximately 22 nucleotides long) – in the bloodstream that has the potential to empower oncologists to not only refine their selection of advanced-stage lung cancer patients who stand to benefit from immunotherapy alone but also identify individuals who may respond more favorably to alternative combinations of therapies.

“By identifying the right therapy, we may be able to increase the life expectancy of these very sick patients,” adds Rajakumar.

The function of miRNAs

Hummingbird chose to focus on small RNAs because “we have found them to be very powerful biomarkers for the detection of cancer,” says Rajakumar. They have a longer half-life in the blood compared to other RNAs, their levels fluctuate, and small changes can have big implications for disease, he adds.

There is more information encoded in an RNA molecule than there may be in a similar DNA molecule, Rajakumar adds. “That is extra information that makes it easier to detect cancer. That's the important and unique aspect of our approach.” 

Recent studies have also shown that signatures of specific miRNAs correlate with treatment outcomes. For example, JTO Clinical Research and Reports published that the miRNA signature miRisk is able to distinguish a group of patients with stage IV NSCLC and high PD-L1 most likely to benefit from adding chemotherapy to their immunotherapy treatment ⁽²⁾. “Chemotherapy can sensitize tumors to immunotherapy, making the treatment more effective,” explains Rajakumar.

Recognizing the potential and driven by a natural fascination with RNA biology, Rajakumar delved deep into the exploration of miRNA molecule identities within the miRisk signature. He has found that they predominantly come from myeloid cells – a type of immune cell.

“That was fascinating,” he says because there are subclasses of myeloid-derived cells involved in cancer suppression. This suggests that the signature miRNAs might be a hallmark of changes in myeloid cells that help make cancer cells more receptive to immunotherapy, he adds. 

A shocking discovery

Once he learned the identities of the miRNAs, he set out to better understand the molecules’ functions. Researchers performed bioinformatic prediction analysis on their five-signature micro-RNAs that make up miRisk and “were shocked” to find that three of them target the PD-L1 gene.

 “We now have this direct functional link between our biomarkers and the drug whose efficacy we're trying to predict,” says Rajakumar.

Rajakumar envisions miRisk as a complementary assay to PD-L1 in advanced-stage lung cancer. Current international treatment guidelines for patients with high PD-L1 recommend giving either immunotherapy alone or immunotherapy and chemotherapy combined. “But there is no clear specification as to which of these options may be better. Individual doctors will use their own judgment to decide.”

“We hope miRisk will enable a better prognosis in all advanced-stage lung cancer patients.”

A clinical validation study of the predictive power of a high miRisk score in advanced-stage lung cancer is scheduled to launch later this year, where it is seeks to show that treatment with a combination of immunotherapy and chemotherapy will be most beneficial. The study will also compare patients with a low miRisk score treated with combination therapy to those treated with immunotherapy alone to understand whether both treatments are similarly effective in this group.

But the research goals don’t end there. Another of Hummingbird’s priority projects is a blood-based lung cancer screening test that could be performed in a doctor’s office. In a study published in the Journal of Thoracic Oncology in July, Hummingbird researchers tested the ability of a signature of 18 specific small RNAs from blood, called miLung, to detect lung cancer early (3).

The prospective study of over 1,300 participants who were candidates for early lung cancer screening found that the performance of miLung was comparable to that of low-dose CT scans. 

Enabling better prognosis with QIAGEN

Hummingbird’s labs rely on QIAGEN’s automated extraction robot, QIAsymphony, to extract small RNAs from QIAGEN’s PAXgene RNA Blood Tubes. This minimizes human error and variation in sample preparation. “The QIAsymphony is the heart and soul of our extraction. That is our workhorse,” Kohlhaas adds. 

For the analysis of small RNA with NGS, Hummingbird adapted and modified QIAGEN’s FastSelect blocking technology that enabled Hummingbird to ignore small RNAs that are abundant but not informative for their test and at the same time, amplify the signals from small RNAs of interest.

“We had an excellent collaboration with QIAGEN,” says Kohlhaas, mentioning the QIAGEN Genomic Services scientists that helped Hummingbird with the FastSelect feasibility studies. “They helped us to win stability in the lab.” Hummingbird scientists have since modified and optimized the blocking technology to suit their specific needs.

Meanwhile, the company is transferring the analysis of their miRisk and miLung signatures from an NGS platform to dPCR using QIAcuity to validate the signatures on an orthogonal platform and to potentially facilitate a kit model test.

“The big advantage of a kit model is that the analysis can be done in a decentralized setting,” Kohlhaas says so that results would be available much quicker.” So far, the company is “impressed” by QIAcuity’s performance, the dPCR platform is able to pick up small RNAs present in minute amounts in a sample.

Looking back, Rajakumar is amazed at what he and the Hummingbird team have accomplished in such a short amount of time.