NSCLC diagnosis 2.0: A strategic view of liquid biopsy

If you run a hospital lab or a diagnostic lab, you are not a stranger to the prevalence, complexity, and recurrence of lung cancer cases. One in five cancer deaths are lung cancer patients, and a staggering 85% of those are non-small-cell lung cancer (NSCLC).^1,2

This mortality rate calls for effective clinical management. This means giving the right drug to the right patient at the right time – which requires real-time tumor monitoring. Meanwhile, the idea of a cancer diagnosis from a blood draw sounded closer to science fiction.

Today, not only is liquid biopsy accessible but also accurate, cost-effective, and patient-friendly.

But the surge in innovation makes it difficult to keep track – let alone have a strategy or foresee the future. What added value does this bring to NSCLC diagnosis? What can today’s patients benefit from? And what are the missing pieces?

Here, we take a closer look at the strategic value and general outlook of liquid biopsy in non-small-cell lung cancer diagnosis and clinical management.

Tissue sampling is still the gold standard for tumor profiling, and CT scans are the norm for detection. However, healthcare management is racing to reap the advantages of liquid biopsy:

• Cost-friendliness
• Non-to-minimal invasiveness with lower complication risk
• Repeatability in case of mistakes or sample insufficiency
• Improved patient experience
• High-quality data
• Boost in diagnostic power and performance
• Lab modernization
• Small sample volumes
• Compared to solid (tissue) biopsy, it provides information from all tumor locations as opposed to a single locus
• Compared to CT scans, it is radiation-free and more sensitive, detecting relapse months ahead

A growing list of tumor components can now be isolated from a blood sample. This list includes well-established components like the two flagship products of liquid biopsy: circulating tumor DNA (ctDNA) and circulating tumor cells (CTCs). It also includes newer, emerging diagnostic targets like platelets, vesicles, and RNA. But despite FDA-cleared applications being either ctDNA or CTCs, the scientific literature is signaling change.

In the past two years, the number of RNA studies has been outgrowing ctDNA, CTCs, and other tumor components – combined. Platelet-derived RNA-Seq was able to distinguish the 55 healthy individuals from the 228 patients with 96% accuracy.³ Two other studies used miRNAs to detect early-stage disease in NSCLC cohorts. The studies showed sensitivities and specificities in the 80-90% ballpark – outperforming ctDNA in the early stages.^4,5 No RNA tests have been cleared for clinical use yet, but this potential future buzzword is something to keep an eye on.

A diagnostic facility can benefit from liquid biopsy in three aspects: early detection, prognosis, and tracking tumor changes.
 

1. Early detection

Nothing holds more promise for the future of cancer diagnosis than a robust early detection system. The ability to catch NSCLC in the controllable stages is key to improving patient outcomes. But how applicable is this?

The first clue we might want to grab is how early tumors can be detected. The answer is: surprisingly early. A 2020 study detected CTCs as early as tumors that are 0.1mm wide (For reference, up to 2cm is still considered a Stage I tumor!) The study used melanoma and breast cancer models and analyzed lymph instead of blood.⁶

This raises a question: could we find similar outcomes in NSCLC, using blood, and capturing other tumor components? The scientific literature answers again:

• A 2018 study used a multi-analyte test called CancerSEEK on a cohort of 1,005 patients of various cancer types. CancerSEEK attempted to detect resectable cancers using ctDNA and protein biomarkers. The scientists at Johns Hopkins were able to detect cancers with an overall 99% specificity and a 69-98% sensitivity.⁷ CancerSEEK is FDA-approved as of 2019.⁸

• Another promising result came up at Stanford University using Cancer Personalized Profiling by deep Sequencing (CAPP-Seq), where 100% of NSCLC tumors in stages II-IV were detected.⁹

Such assuring results increased interest (and endorsements) in early detection technology. Emerging new technologies promise to circumvent pre-analytical hurdles. Electric field-induced release and measurement – or EFIRM – strikingly detected two major lung tumor mutations from the saliva of NSCLC patients with little to no sample preparation procedure.¹⁰

Another example is a new ‘fragmentomics’ approach. Fragmentomics discern ctDNA from healthy cfDNA using its differential fragmentation pattern, this new technology promises to reduce the cost of ctDNA analysis compared to NGS by nearly ten-fold.¹¹ Although in some cases, Stage I NSCLC could be invisible to plasma analysis. But a closer look can say something: false negatives showed significantly better prognosis and survival compared to those who showed disease signals early.¹²

Theoretically, this means that when liquid biopsy falls short in early detection, tumor components would still show prognostic value.
 

2. Prognosis

In the majority of NSCLC cases, physicians encounter a multitude of drug resistance mechanisms and relapses, making non-small-cell lung cancer prognosis difficult.¹³ This leaves decision-makers with an overwhelming number of options and combinations of therapeutics to consider during treatment. Being able to measure predictive and prognostic parameters alleviates the stress of decision-making.

Liquid biopsy proved helpful in collecting data to establish the best therapeutic options. Here are two examples:

• Using ctDNA analysis, the EURTAC trial intended to compare the efficiency of erlotinib to chemotherapy as a first-line treatment in NSCLC. By using PCR, the trial pointed out that the presence of the L858R EGFR mutation marked a reduction of overall survival by half (13.7 vs. 27.7 months).¹⁴

• In another cohort of NSCLC patients, the presence of fewer than 5 CTCs per blood sample correlated to a staggering two-to-three-fold increase in progression-free and overall survival.¹⁵

Smaller tumor components carry a predictive value of their own. Exosome-derived miRNA has been proven to induce resistance to cisplatin (DDP) in A549 cell lines (a lung adenocarcinoma cell line). A different miRNA molecule was shown to desensitize cells to the same chemotherapy.¹⁶ Extrapolation of such in vitro studies on patient cohorts could prove valuable for non-small-cell lung cancer prognosis.
 

3. Tracking changes in tumor profile

In short, this helps follow up lung cancer with new treatments. Longitudinal follow-up and monitoring minimal residual disease (MRD) are some of the most widely – and rapidly – adopted uses of liquid biopsy, especially in lung cancer. The patient’s plasma monitoring allows for baseline measurement of the tumor's mutational status. This occurs during NSCLC diagnosis and before any treatment is administered.

Screening for EGFR mutations is one approved use of NGS which select patients for first-line TKI treatment. Initial screening allows for the selection of drugs according to the tumor profile provided by NGS. However, this does not guarantee a relapse-free journey.

As mentioned earlier, NSCLC tumors switch between resistance mechanisms to survive the pressure of the new treatment, and relapse is almost inevitable in NSCLC.¹³ This transformation renders the first line of treatment inefficient. A successful selection of the second-line treatment warrants tracking mutations and anticipating resistance mechanisms early by a repetitive follow-up of the tumor profile.

Plasma analysis is now the preferred choice for clinicians for longitudinal follow-up of patients. Not only does it anticipate relapse but also depicts the tumor profile and which mutations will amplify after the first line of treatment.

Without a doubt, liquid biopsy clears the fog and narrows down imprecision on different fronts. The field promises to cover tumor biology blind spots, enhance patient outcomes, and improve the quality of life for cancer patients.

A liquid biopsy might look close to dethroning its tissue-based counterpart as the gold standard – but we are not quite there yet.

According to the director of the Thoracic Medical Oncology and the Early Clinical Trials Departments at the University of Maryland, Professor Christian Rolfo, there are two major obstacles to adoption. First is the price of liquid biopsy – especially for NGS assays that fall in the $1,000-$7,000 ballpark per run. The second big obstacle is false negative results for non-shedding patients, despite such assays being awaited eagerly. 

Using liquid biopsy still does not guarantee that a patient will receive matching therapy. Additionally, the technology favors specificity over sensitivity. Meaning that it is more likely to have a false negative than a false positive result. This means that a false negative should be followed up with a tissue biopsy whenever feasible.

CTCs have a predominant prognostic value. However, enrichment is a drawback. If the prognostic value is tied to the number of CTCs isolated, then it is not reproducible because other techniques might enrich a different number of CTCs for the same case. Larger studies with standardized operations are needed to establish such relationships with confidence.

What can healthcare leaders do? With clinical uses still limited to a few specific cases, the current state of the art warrants improvements in the following aspects:

1. Technology transfer: new technology does not spin off the bench straight to the clinic. Technology transfer is needed to mold the many aspects of liquid biopsy into a more practical form for day-to-day clinical use. Tech transfer ensures a user-friendly interface and reproducibility of results. Expertise in this area will be needed in the near future

2. Standardization of procedures: a French study found 5 high-risk patients with COPD using ISET for CTC capture.^17,18 A later prospective study in France failed to confirm the results and detect the CTCs in 13 of 15 patients with lung cancer.¹⁹ This is one of the countless examples that highlight the need for standardization of procedures to ensure consistency and reproducibility between platforms

3. Clinical validation: early detection, timely treatment, and spotting of tumor changes are necessary. But how do we know that those feats translate to numbers? Clinical studies using large enough cohorts and providing enough data volume statistically demonstrate the added value of liquid biopsy

4. Artificial intelligence: introducing A.I. to NSCLC diagnosis will be something to watch. Machine learning will leverage the different steps of data storage, retrieval, and analysis. It can also boost decision-making by narrowing down therapy options as well as present patterns of previous trials, giving clinicians a landscape of scenarios of what to expect after every decision

5. Academic research: fundamental research will be valuable in layering the rationale and credibility for using liquid biopsy. Many questions about the biology of tumors, communication, and shedding remain unanswered. Academia will be a catalyst in answering ambiguities and fueling the journey to the clinic

A few years ago, the literature presented data with a dose of constraint. Today, we sense a more optimistic tone. This trend means that liquid biopsy is shaping up. More answers are still needed. But healthcare leaders and lab managers can do more than wait.

Frontline healthcare personnel can collaborate with researchers on clinical studies. This allows patients to experience the technology and provides data that feeds into the endorsement course. In the meantime, the healthy path is not to abuse or overhype this paradigm. Liquid biopsy should rather be juxtaposed with standard diagnostic tools, like imaging and physical examinations.

Liquid biopsy is here to stay, and an accelerated endorsement is expected, but the journey to the clinic is far from over.